feat: redirect / to /dashboard

- Remove unused imports: time/asyncio/Path/Optional/re/Counter/jieba
  across backfill_embeddings, bucket_manager, embedding_engine, import_memory
- Drop unused local var `old_primary` in reclassify_domains
- Fix 3 placeholder-less f-strings in backfill_embeddings, migrate_to_domains,
  reclassify_domains
- Delete legacy quick-check scripts test_smoke.py / test_tools.py (replaced
  by tests/ pytest suite, never collected by pytest)
- Delete backup_20260405_2124/ (stale local snapshot, no code references)

Verified: pyflakes *.py → 0 warnings; pytest tests/ → 237 passed, 7 skipped.

.gitignore

@@ -15,3 +15,7 @@ scarp_paper
 backup_*/
 *.db
 import_state.json
+data/
+tests/integration/
+tests/regression/
+tests/unit/

BEHAVIOR_SPEC.md

@@ -0,0 +1,632 @@
+# Ombre Brain 用户全流程行为规格书
+
+> 版本：基于 server.py / bucket_manager.py / decay_engine.py / dehydrator.py / embedding_engine.py / CLAUDE_PROMPT.md / config.example.yaml
+
+---
+
+## 一、系统角色说明
+
+### 1.1 参与方总览
+
+| 角色 | 实体 | 职责边界 |
+|------|------|---------|
+| **用户** | 人类 | 发起对话，提供原始内容；可直接访问 Dashboard Web UI |
+| **Claude（模型端）** | LLM（如 Claude 3.x）| 理解语义、决策何时调用工具、用自然语言回应用户；不直接操作文件 |
+| **OB 服务端** | `server.py` + 各模块 | 接收 MCP 工具调用，执行持久化、搜索、衰减；对 Claude 不透明 |
+
+### 1.2 Claude 端职责边界
+- **必须做**：每次新对话第一步无参调用 `breath()`；对话内容有记忆价值时主动调用 `hold` / `grow`
+- **不做**：不直接读写 `.md` 文件；不执行衰减计算；不操作 SQLite
+- **决策权**：Claude 决定是否存、存哪些、何时 resolve；OB 决定如何存（合并/新建）
+
+### 1.3 OB 服务端内部模块职责
+
+| 模块 | 核心职责 |
+|------|---------|
+| `server.py` | 注册 MCP 工具（`breath/hold/grow/trace/pulse/dream`）；路由 Dashboard HTTP 请求；`_merge_or_create()` 合并逻辑中枢 |
+| `bucket_manager.py` | 桶 CRUD；多维搜索（fuzzy + embedding 双通道）；`touch()` 激活刷新；`_time_ripple()` 时间波纹 |
+| `dehydrator.py` | `analyze()` 自动打标；`merge()` 内容融合；`digest()` 日记拆分；`dehydrate()` 内容压缩 |
+| `embedding_engine.py` | `generate_and_store()` 生成向量并存 SQLite；`search_similar()` 余弦相似度检索 |
+| `decay_engine.py` | `calculate_score()` 衰减分计算；`run_decay_cycle()` 周期扫描归档；后台定时循环 |
+| `utils.py` | 配置加载；路径安全校验；ID 生成；token 估算 |
+
+---
+
+## 二、场景全流程
+
+---
+
+### 场景 1：新对话开始（冷启动，无历史记忆）
+
+**用户操作**：打开新对话窗口，说第一句话
+
+**Claude 行为**：在任何回复之前，先调用 `breath()`（无参）
+
+**OB 工具调用**：
+```
+breath(query="", max_tokens=10000, domain="", valence=-1, arousal=-1, max_results=20, importance_min=-1)
+```
+
+**系统内部发生什么**：
+1. `decay_engine.ensure_started()` — 懒加载启动后台衰减循环（若未运行）
+2. 进入"浮现模式"（`not query or not query.strip()`）
+3. `bucket_mgr.list_all(include_archive=False)` — 遍历 `permanent/` + `dynamic/` + `feel/` 目录，加载所有 `.md` 文件的 frontmatter + 正文
+4. 筛选钉选桶（`pinned=True` 或 `protected=True`）
+5. 筛选未解决桶（`resolved=False`，排除 `permanent/feel/pinned`）
+6. **冷启动检测**：找 `activation_count==0 && importance>=8` 的桶，最多取 2 个插入排序最前（**决策：`create()` 初始化应为 0，区分"创建"与"被主动召回"，见 B-04**）
+7. 按 `decay_engine.calculate_score(metadata)` 降序排列剩余未解决桶
+8. 对 top-20 以外随机洗牌（top-1 固定，2~20 随机）
+9. 截断到 `max_results` 条
+10. 对每个桶调用 `dehydrator.dehydrate(strip_wikilinks(content), clean_meta)` 压缩摘要
+11. 按 `max_tokens` 预算截断输出
+
+**返回结果**：
+- 无记忆时：`"权重池平静，没有需要处理的记忆。"`
+- 有记忆时：`"=== 核心准则 ===\n📌 ...\n\n=== 浮现记忆 ===\n[权重:X.XX] [bucket_id:xxx] ..."`
+
+**注意**：浮现模式**不调用** `touch()`，不重置衰减计时器
+
+---
+
+### 场景 2：新对话开始（有历史记忆，breath 自动浮现）
+
+（与场景 1 相同流程，区别在于桶文件已存在）
+
+**Claude 行为（完整对话启动序列，来自 CLAUDE_PROMPT.md）**：
+```
+1. breath()               — 浮现未解决记忆
+2. dream()                — 消化最近记忆，有沉淀写 feel
+3. breath(domain="feel")  — 读取之前的 feel
+4. 开始和用户说话
+```
+
+**`breath(domain="feel")` 内部流程**：
+1. 检测到 `domain.strip().lower() == "feel"` → 进入 feel 专用通道
+2. `bucket_mgr.list_all()` 过滤 `type=="feel"` 的桶
+3. 按 `created` 降序排列
+4. 按 `max_tokens` 截断，不压缩（直接展示原文）
+5. 返回：`"=== 你留下的 feel ===\n[时间] [bucket_id:xxx]\n内容..."`
+
+---
+
+### 场景 3：用户说了一件事，Claude 决定存入记忆（hold）
+
+**用户操作**：例如"我刚刚拿到了实习 offer，有点激动"
+
+**Claude 行为**：判断值得记忆，调用：
+```python
+hold(content="用户拿到实习 offer，情绪激动", importance=7)
+```
+
+**OB 工具调用**：`hold(content, tags="", importance=7, pinned=False, feel=False, source_bucket="", valence=-1, arousal=-1)`
+
+**系统内部发生什么**：
+
+1. `decay_engine.ensure_started()`
+2. 输入校验：`content.strip()` 非空
+3. `importance = max(1, min(10, 7))` = 7
+4. `extra_tags = []`（未传 tags）
+5. **自动打标**：`dehydrator.analyze(content)` → 调用 `_api_analyze()` → LLM 返回 JSON
+   - 返回示例：`{"domain": ["成长", "求职"], "valence": 0.8, "arousal": 0.7, "tags": ["实习", "offer", "激动", ...], "suggested_name": "实习offer获得"}`
+   - 失败时降级：`{"domain": ["未分类"], "valence": 0.5, "arousal": 0.3, "tags": [], "suggested_name": ""}`
+6. 合并 `auto_tags + extra_tags` 去重
+7. **合并检测**：`_merge_or_create(content, tags, importance=7, domain, valence, arousal, name)`
+   - `bucket_mgr.search(content, limit=1, domain_filter=domain)` — 搜索最相似的桶
+   - 若最高分 > `config["merge_threshold"]`（默认 75）且该桶非 pinned/protected：
+     - `dehydrator.merge(old_content, new_content)` → `_api_merge()` → LLM 融合
+     - `bucket_mgr.update(bucket_id, content=merged, tags=union, importance=max, domain=union, valence=avg, arousal=avg)`
+     - `embedding_engine.generate_and_store(bucket_id, merged_content)` 更新向量
+     - 返回 `(bucket_name, True)`
+   - 否则：
+     - `bucket_mgr.create(content, tags, importance=7, domain, valence, arousal, name)` → 写 `.md` 文件到 `dynamic/<主题域>/` 目录
+     - `embedding_engine.generate_and_store(bucket_id, content)` 生成并存储向量
+     - 返回 `(bucket_id, False)`
+
+**返回结果**：
+- 新建：`"新建→实习offer获得 成长,求职"`
+- 合并：`"合并→求职经历 成长,求职"`
+
+**bucket_mgr.create() 详情**：
+- `generate_bucket_id()` → `uuid4().hex[:12]`
+- `sanitize_name(name)` → 正则清洗，最长 80 字符
+- 写 YAML frontmatter + 正文到 `safe_path(domain_dir, f"{name}_{id}.md")`
+- frontmatter 字段：`id, name, tags, domain, valence, arousal, importance, type, created, last_active, activation_count=0`（**决策：初始为 0，`touch()` 首次被召回后变为 1**）
+
+---
+
+### 场景 4：用户说了一段长日记，Claude 整理存入（grow）
+
+**用户操作**：发送一大段混合内容，如"今天去医院体检，结果还好；晚上和朋友吃饭聊了很多；最近有点焦虑..."
+
+**Claude 行为**：
+```python
+grow(content="今天去医院体检，结果还好；晚上和朋友吃饭聊了很多；最近有点焦虑...")
+```
+
+**系统内部发生什么**：
+
+1. `decay_engine.ensure_started()`
+2. 内容长度检查：`len(content.strip()) < 30` → 若短于 30 字符走**快速路径**（`dehydrator.analyze()` + `_merge_or_create()`，跳过 digest）
+3. **日记拆分**（正常路径）：`dehydrator.digest(content)` → `_api_digest()` → LLM 调用 `DIGEST_PROMPT`
+   - LLM 返回 JSON 数组，每项含：`name, content, domain, valence, arousal, tags, importance`
+   - `_parse_digest()` 安全解析，校验 valence/arousal 范围
+4. 对每个 `item` 调用 `_merge_or_create(item["content"], item["tags"], item["importance"], item["domain"], item["valence"], item["arousal"], item["name"])`
+   - 每项独立走合并或新建逻辑（同场景 3）
+   - 单条失败不影响其他条（`try/except` 隔离）
+
+**返回结果**：
+```
+3条|新2合1
+📝体检结果
+📌朋友聚餐
+📎近期焦虑情绪
+```
+
+---
+
+### 场景 5：用户想找某段记忆（breath 带 query 检索）
+
+**用户操作**：例如"还记得我之前说过关于实习的事吗"
+
+**Claude 行为**：
+```python
+breath(query="实习", domain="成长", valence=0.7, arousal=0.5)
+```
+
+**系统内部发生什么**：
+
+1. `decay_engine.ensure_started()`
+2. 检测到 `query` 非空，进入**检索模式**
+3. 解析 `domain_filter = ["成长"]`，`q_valence=0.7`，`q_arousal=0.5`
+4. **关键词检索**：`bucket_mgr.search(query, limit=20, domain_filter, q_valence, q_arousal)`
+   - **Layer 1**：domain 预筛 → 仅保留 domain 包含"成长"的桶；若为空则回退全量
+   - **Layer 1.5**（embedding 已开启时）：`embedding_engine.search_similar(query, top_k=50)` → 用 embedding 候选集替换/缩小精排范围
+   - **Layer 2**：多维加权精排：
+     - `_calc_topic_score()`: `fuzz.partial_ratio(query, name)×3 + domain×2.5 + tags×2 + body×1`，归一化 0~1
+     - `_calc_emotion_score()`: `1 - √((v差²+a差²)/2)`，0~1
+     - `_calc_time_score()`: `e^(-0.02×days_since_last_active)`，0~1
+     - `importance_score`: `importance / 10`
+     - `total = topic×4 + emotion×2 + time×1.5 + importance×1`，归一化到 0~100
+     - 过滤 `score >= fuzzy_threshold`（默认 50）
+     - 通过阈值后，`resolved` 桶仅在排序时降权 ×0.3（不影响是否被检出）
+     - 返回最多 `limit` 条
+5. 排除 pinned/protected 桶（它们在浮现模式展示）
+6. **向量补充通道**（server.py 额外层）：`embedding_engine.search_similar(query, top_k=20)` → 相似度 > 0.5 的桶补充到结果集（标记 `vector_match=True`）
+7. 对每个结果：
+   - 记忆重构：若传了 `q_valence`，展示层 valence 做微调：`shift = (q_valence - 0.5) × 0.2`，最大 ±0.1
+   - `dehydrator.dehydrate(strip_wikilinks(content), clean_meta)` 压缩摘要
+   - `bucket_mgr.touch(bucket_id)` — 刷新 `last_active` + `activation_count += 1` + 触发 `_time_ripple()`（对 48h 内创建的邻近桶 activation_count + 0.3，最多 5 个桶）
+8. **随机漂流**：若检索结果 < 3 且 `random.random() < 0.4`，随机从 `decay_score < 2.0` 的旧桶里取 1~3 条，标注 `[surface_type: random]`
+
+**返回结果**：
+```
+[bucket_id:abc123] [重要度:7] [主题:成长] 实习offer获得：...
+[语义关联] [bucket_id:def456] 求职经历...
+--- 忽然想起来 ---
+[surface_type: random] 某段旧记忆...
+```
+
+---
+
+### 场景 6：用户想查看所有记忆状态（pulse）
+
+**用户操作**："帮我看看你现在都记得什么"
+
+**Claude 行为**：
+```python
+pulse(include_archive=False)
+```
+
+**系统内部发生什么**：
+
+1. `bucket_mgr.get_stats()` — 遍历三个目录，统计文件数量和 KB 大小
+2. `bucket_mgr.list_all(include_archive=False)` — 加载全部桶
+3. 对每个桶：`decay_engine.calculate_score(metadata)` 计算当前权重分
+4. 按类型/状态分配图标：📌钉选 / 📦permanent / 🫧feel / 🗄️archived / ✅resolved / 💭普通
+5. 拼接每桶摘要行：`名称 bucket_id 主题 情感坐标 重要度 权重 标签`
+
+**返回结果**：
+```
+=== Ombre Brain 记忆系统 ===
+固化记忆桶: 2 个
+动态记忆桶: 15 个
+归档记忆桶: 3 个
+总存储大小: 48.3 KB
+衰减引擎: 运行中
+
+=== 记忆列表 ===
+📌 [核心原则] bucket_id:abc123 主题:内心 情感:V0.8/A0.5 ...
+💭 [实习offer获得] bucket_id:def456 主题:成长 情感:V0.8/A0.7 ...
+```
+
+---
+
+### 场景 7：用户想修改/标记已解决/删除某条记忆（trace）
+
+#### 7a 标记已解决
+
+**Claude 行为**：
+```python
+trace(bucket_id="abc123", resolved=1)
+```
+
+**系统内部**：
+1. `resolved in (0, 1)` → `updates["resolved"] = True`
+2. `bucket_mgr.update("abc123", resolved=True)` → 读取 `.md` 文件，更新 frontmatter 中 `resolved=True`，写回，**桶留在原 `dynamic/` 目录，不移动**
+3. 后续 `breath()` 浮现时：该桶 `decay_engine.calculate_score()` 乘以 `resolved_factor=0.05`（若同时 `digested=True` 则 ×0.02），自然降权，最终由 decay 引擎在得分 < threshold 时归档
+4. `bucket_mgr.search()` 中该桶得分乘以 0.3 降权，但仍可被关键词激活
+
+> ⚠️ **代码 Bug B-01**：当前实现中 `update(resolved=True)` 会将桶**立即移入 `archive/`**，导致桶完全消失于所有搜索路径，与上述规格不符。需移除 `bucket_manager.py` `update()` 中 resolved → `_move_bucket(archive_dir)` 的自动归档逻辑。
+
+**返回**：`"已修改记忆桶 abc123: resolved=True → 已沉底，只在关键词触发时重新浮现"`
+
+#### 7b 修改元数据
+
+```python
+trace(bucket_id="abc123", name="新名字", importance=8, tags="焦虑,成长")
+```
+
+**系统内部**：收集非默认值字段 → `bucket_mgr.update()` 批量更新 frontmatter
+
+#### 7c 删除
+
+```python
+trace(bucket_id="abc123", delete=True)
+```
+
+**系统内部**：
+1. `bucket_mgr.delete("abc123")` → `_find_bucket_file()` 定位文件 → `os.remove(file_path)`
+2. `embedding_engine.delete_embedding("abc123")` → SQLite `DELETE WHERE bucket_id=?`
+3. 返回：`"已遗忘记忆桶: abc123"`
+
+---
+
+### 场景 8：记忆长期未被激活，自动衰减归档（后台 decay）
+
+**触发方式**：服务启动后，`decay_engine.start()` 创建后台 asyncio Task，每 `check_interval_hours`（默认 24h）执行一次 `run_decay_cycle()`
+
+**系统内部发生什么**：
+
+1. `bucket_mgr.list_all(include_archive=False)` — 获取所有活跃桶
+2. 跳过 `type in ("permanent","feel")` 或 `pinned=True` 或 `protected=True` 的桶
+3. **自动 resolve**：若 `importance <= 4` 且距上次激活 > 30 天且 `resolved=False` → `bucket_mgr.update(bucket_id, resolved=True)`
+4. 对每桶调用 `calculate_score(metadata)`：
+
+   **短期（days_since ≤ 3）**：
+   ```
+   time_weight = 1.0 + e^(-hours/36)  (t=0→×2.0, t=36h→×1.5)
+   emotion_weight = base(1.0) + arousal × arousal_boost(0.8)
+   combined = time_weight×0.7 + emotion_weight×0.3
+   base_score = importance × activation_count^0.3 × e^(-λ×days) × combined
+   ```
+   **长期（days_since > 3）**：
+   ```
+   combined = emotion_weight×0.7 + time_weight×0.3
+   ```
+   **修正因子**：
+   - `resolved=True` → ×0.05
+   - `resolved=True && digested=True` → ×0.02
+   - `arousal > 0.7 && resolved=False` → ×1.5（高唤醒紧迫加成）
+   - `pinned/protected/permanent` → 返回 999.0（永不衰减）
+   - `type=="feel"` → 返回 50.0（固定）
+
+5. `score < threshold`（默认 0.3）→ `bucket_mgr.archive(bucket_id)` → `_move_bucket()` 将文件从 `dynamic/` 移动到 `archive/` 目录，更新 frontmatter `type="archived"`
+
+**返回 stats**：`{"checked": N, "archived": N, "auto_resolved": N, "lowest_score": X}`
+
+---
+
+### 场景 9：用户使用 dream 工具进行记忆沉淀
+
+**触发**：Claude 在对话启动时，`breath()` 之后调用 `dream()`
+
+**OB 工具调用**：`dream()`（无参数）
+
+**系统内部发生什么**：
+
+1. `bucket_mgr.list_all()` → 过滤非 `permanent/feel/pinned/protected` 桶
+2. 按 `created` 降序取前 10 条（最近新增的记忆）
+3. 对每条拼接：名称、resolved 状态、domain、V/A、创建时间、正文前 500 字符
+4. **连接提示**（embedding 已开启 && 桶数 >= 2）：
+   - 取每个最近桶的 embedding（`embedding_engine.get_embedding(bucket_id)`）
+   - 两两计算 `_cosine_similarity()`，找相似度最高的对
+   - 若 `best_sim > 0.5` → 输出提示：`"[名A] 和 [名B] 似乎有关联 (相似度:X.XX)"`
+5. **feel 结晶提示**（embedding 已开启 && feel 数 >= 3）：
+   - 对所有 feel 桶两两计算相似度
+   - 若某 feel 与 >= 2 个其他 feel 相似度 > 0.7 → 提示升级为 pinned 桶
+6. 返回标准 header 说明（引导 Claude 自省）+ 记忆列表 + 连接提示 + 结晶提示
+
+**Claude 后续行为**（根据 CLAUDE_PROMPT 引导）：
+- `trace(bucket_id, resolved=1)` 放下可以放下的
+- `hold(content="...", feel=True, source_bucket="xxx", valence=0.6)` 写感受
+- 无沉淀则不操作
+
+---
+
+### 场景 10：用户使用 feel 工具记录 Claude 的感受
+
+**触发**：Claude 在 dream 后决定记录某段记忆带来的感受
+
+**OB 工具调用**：
+```python
+hold(content="她问起了警校的事，我感觉她在用问题保护自己，问是为了不去碰那个真实的恐惧。", feel=True, source_bucket="abc123", valence=0.45, arousal=0.4)
+```
+
+**系统内部发生什么**：
+
+1. `feel=True` → 进入 feel 专用路径，跳过自动打标和合并检测
+2. `feel_valence = valence`（Claude 自身视角的情绪，非事件情绪）
+3. `bucket_mgr.create(content, tags=[], importance=5, domain=[], valence=feel_valence, arousal=feel_arousal, bucket_type="feel")` → 写入 `feel/` 目录
+4. `embedding_engine.generate_and_store(bucket_id, content)` — feel 桶同样有向量（供 dream 结晶检测使用）
+5. 若 `source_bucket` 非空：`bucket_mgr.update(source_bucket, digested=True, model_valence=feel_valence)` → 标记源记忆已消化
+   - 此后该源桶 `calculate_score()` 中 `resolved_factor = 0.02`（accelerated fade）
+
+**衰减特性**：feel 桶 `type=="feel"` → `calculate_score()` 固定返回 50.0，永不归档
+**检索特性**：不参与普通 `breath()` 浮现；只通过 `breath(domain="feel")` 读取
+
+**返回**：`"🫧feel→<bucket_id>"`
+
+---
+
+### 场景 11：用户带 importance_min 参数批量拉取重要记忆
+
+**Claude 行为**：
+```python
+breath(importance_min=8)
+```
+
+**系统内部发生什么**：
+
+1. `importance_min >= 1` → 进入**批量拉取模式**，完全跳过语义搜索
+2. `bucket_mgr.list_all(include_archive=False)` 全量加载
+3. 过滤 `importance >= 8` 且 `type != "feel"` 的桶
+4. 按 `importance` 降序排列，截断到最多 20 条
+5. 对每条调用 `dehydrator.dehydrate()` 压缩，按 `max_tokens`（默认 10000）预算截断
+
+**返回**：
+```
+[importance:10] [bucket_id:xxx] ...（核心原则）
+---
+[importance:9] [bucket_id:yyy] ...
+---
+[importance:8] [bucket_id:zzz] ...
+```
+
+---
+
+### 场景 12：embedding 向量化检索场景（开启 embedding 时）
+
+**前提**：`config.yaml` 中 `embedding.enabled: true` 且 `OMBRE_API_KEY` 已配置
+
+**embedding 介入的两个层次**：
+
+#### 层次 A：BucketManager.search() 内的 Layer 1.5 预筛
+- 调用点：`bucket_mgr.search()` → Layer 1.5
+- 函数：`embedding_engine.search_similar(query, top_k=50)` → 生成查询 embedding → SQLite 全量余弦计算 → 返回 `[(bucket_id, similarity)]` 按相似度降序
+- 作用：将精排候选集从所有桶缩小到向量最近邻的 50 个，加速后续多维精排
+
+#### 层次 B：server.py breath 的额外向量通道
+- 调用点：`breath()` 检索模式中，keyword 搜索完成后
+- 函数：`embedding_engine.search_similar(query, top_k=20)` → 相似度 > 0.5 的桶补充到结果集
+- 标注：补充桶带 `[语义关联]` 前缀
+
+**向量存储路径**：
+- 新建桶后：`embedding_engine.generate_and_store(bucket_id, content)` → `_generate_embedding(text[:2000])` → API 调用 → `_store_embedding()` → SQLite `INSERT OR REPLACE`
+- 合并更新后：同上，用 merged content 重新生成
+- 删除桶时：`embedding_engine.delete_embedding(bucket_id)` → `DELETE FROM embeddings`
+
+**SQLite 结构**：
+```sql
+CREATE TABLE embeddings (
+    bucket_id TEXT PRIMARY KEY,
+    embedding TEXT NOT NULL,   -- JSON 序列化的 float 数组
+    updated_at TEXT NOT NULL
+)
+```
+
+**相似度计算**：`_cosine_similarity(a, b)` = dot(a,b) / (|a| × |b|)
+
+---
+
+## 三、边界与降级行为
+
+| 场景 | 异常情况 | 降级行为 |
+|------|---------|---------|
+| `breath()` 浮现 | 桶目录为空 | 返回 `"权重池平静，没有需要处理的记忆。"` |
+| `breath()` 浮现 | `list_all()` 异常 | 返回 `"记忆系统暂时无法访问。"` |
+| `breath()` 检索 | `bucket_mgr.search()` 异常 | 返回 `"检索过程出错，请稍后重试。"` |
+| `breath()` 检索 | embedding 不可用 / API 失败 | `logger.warning()` 记录，跳过向量通道，仅用 keyword 检索 |
+| `breath()` 检索 | 结果 < 3 条 | 40% 概率从低权重旧桶随机浮现 1~3 条，标注 `[surface_type: random]` |
+| `hold()` 自动打标 | `dehydrator.analyze()` 失败 | 降级到默认值：`domain=["未分类"], valence=0.5, arousal=0.3, tags=[], name=""` |
+| `hold()` 合并检测 | `bucket_mgr.search()` 失败 | `logger.warning()`，直接走新建路径 |
+| `hold()` 合并 | `dehydrator.merge()` 失败 | `logger.warning()`，跳过合并，直接新建 |
+| `hold()` embedding | API 失败 | `try/except` 吞掉，embedding 缺失但不影响存储 |
+| `grow()` 日记拆分 | `dehydrator.digest()` 失败 | 返回 `"日记整理失败: {e}"` |
+| `grow()` 单条处理失败 | 单个 item 异常 | `logger.warning()` + 标注 `⚠️条目名`，其他条目正常继续 |
+| `grow()` 内容 < 30 字 | — | 快速路径：`analyze()` + `_merge_or_create()`，跳过 `digest()`（节省 token） |
+| `trace()` | `bucket_mgr.get()` 返回 None | 返回 `"未找到记忆桶: {bucket_id}"` |
+| `trace()` | 未传任何可修改字段 | 返回 `"没有任何字段需要修改。"` |
+| `pulse()` | `get_stats()` 失败 | 返回 `"获取系统状态失败: {e}"` |
+| `dream()` | embedding 未开启 | 跳过连接提示和结晶提示，仅返回记忆列表 |
+| `dream()` | 桶列表为空 | 返回 `"没有需要消化的新记忆。"` |
+| `decay_cycle` | `list_all()` 失败 | 返回 `{"checked":0, "archived":0, ..., "error": str(e)}`，不终止后台循环 |
+| `decay_cycle` | 单桶 `calculate_score()` 失败 | `logger.warning()`，跳过该桶继续 |
+| 所有 feel 操作 | `source_bucket` 不存在 | `logger.warning()` 记录，feel 桶本身仍成功创建 |
+| `dehydrator.dehydrate()` / `analyze()` / `merge()` / `digest()` | API 不可用（`api_available=False`）| **直接向 MCP 调用端明确报错（`RuntimeError`）**，无本地降级。本地关键词提取质量不足以替代语义打标与合并，静默降级比报错更危险（可能产生错误分类记忆）。 |
+| `embedding_engine.search_similar()` | `enabled=False` | 直接返回 `[]`，调用方 fallback 到 keyword 搜索 |
+
+---
+
+## 四、数据流图
+
+### 4.1 一条记忆的完整生命周期
+
+```
+用户输入内容
+     │
+     ▼
+Claude 决策: hold / grow / 自动
+     │
+     ├─[grow 长内容]──→ dehydrator.digest(content)
+     │                    DIGEST_PROMPT → LLM API
+     │                    返回 [{name,content,domain,...}]
+     │                    ↓ 每条独立处理 ↓
+     │
+     └─[hold 单条]──→ dehydrator.analyze(content)
+                          ANALYZE_PROMPT → LLM API
+                          返回 {domain, valence, arousal, tags, suggested_name}
+                          │
+                          ▼
+                    _merge_or_create()
+                          │
+                    bucket_mgr.search(content, limit=1, domain_filter)
+                          │
+                    ┌─────┴─────────────────────────┐
+                    │ score > merge_threshold (75)?  │
+                    │                                │
+                   YES                               NO
+                    │                                │
+                    ▼                                ▼
+           dehydrator.merge(              bucket_mgr.create(
+             old_content, new)              content, tags,
+             MERGE_PROMPT → LLM             importance, domain,
+                    │                       valence, arousal,
+                    ▼                       bucket_type="dynamic"
+           bucket_mgr.update(...)         )
+                    │                        │
+                    └──────────┬─────────────┘
+                               │
+                               ▼
+                    embedding_engine.generate_and_store(
+                      bucket_id, content)
+                      → _generate_embedding(text[:2000])
+                      → API 调用 (gemini-embedding-001)
+                      → _store_embedding() → SQLite
+                               │
+                               ▼
+                    文件写入: {buckets_dir}/dynamic/{domain}/{name}_{id}.md
+                    YAML frontmatter:
+                      id, name, tags, domain, valence, arousal,
+                      importance, type="dynamic", created, last_active,
+                      activation_count=0   # B-04: starts at 0; touch() bumps to 1+
+                               │
+                               ▼
+          ┌─────── 记忆桶存活期 ──────────────────────────────────────┐
+          │                                                           │
+          │  每次被 breath(query) 检索命中:                           │
+          │    bucket_mgr.touch(bucket_id)                           │
+          │      → last_active = now_iso()                           │
+          │      → activation_count += 1                            │
+          │      → _time_ripple(source_id, now, hours=48)           │
+          │        对 48h 内邻近桶 activation_count += 0.3           │
+          │                                                           │
+          │  被 dream() 消化:                                        │
+          │    hold(feel=True, source_bucket=id) →                  │
+          │    bucket_mgr.update(id, digested=True)                 │
+          │                                                           │
+          │  被 trace(resolved=1) 标记:                              │
+          │    resolved=True → decay score ×0.05 (或 ×0.02)        │
+          │                                                           │
+          └───────────────────────────────────────────────────────────┘
+                               │
+                               ▼
+          decay_engine 后台循环 (每 check_interval_hours=24h)
+            run_decay_cycle()
+              → 列出所有动态桶
+              → calculate_score(metadata)
+                  importance × activation_count^0.3
+                  × e^(-λ×days)
+                  × combined_weight
+                  × resolved_factor
+                  × urgency_boost
+              → score < threshold (0.3)?
+                               │
+                         ┌─────┴──────┐
+                         │            │
+                        YES           NO
+                         │            │
+                         ▼            ▼
+              bucket_mgr.archive(id)  继续存活
+                → _move_bucket()
+                → 文件移动到 archive/
+                → frontmatter type="archived"
+                         │
+                         ▼
+             记忆桶归档（不再参与浮现/搜索）
+             但文件仍存在，可通过 pulse(include_archive=True) 查看
+```
+
+### 4.2 feel 桶的特殊路径
+
+```
+hold(feel=True, source_bucket="xxx", valence=0.45)
+         │
+         ▼
+  bucket_mgr.create(bucket_type="feel")
+  写入 feel/ 目录
+         │
+         ├─→ embedding_engine.generate_and_store()（供 dream 结晶检测）
+         │
+         └─→ bucket_mgr.update(source_bucket, digested=True, model_valence=0.45)
+                    源桶 resolved_factor → 0.02
+                    加速衰减直到归档
+
+feel 桶自身:
+  - calculate_score() 返回固定 50.0
+  - 不参与普通 breath 浮现
+  - 不参与 dreaming 候选
+  - 只通过 breath(domain="feel") 读取
+  - 永不归档
+```
+
+---
+
+## 五、代码与规格差异汇总（审查版）
+
+> 本节由完整源码审查生成（2026-04-21），记录原待实现项最终状态、新发现 Bug 及参数决策。
+
+---
+
+### 5.1 原待实现项最终状态
+
+| 编号 | 原描述 | 状态 | 结论 |
+|------|--------|------|------|
+| ⚠️-1 | `dehydrate()` 无本地降级 fallback | **已确认为设计决策** | API 不可用时直接向 MCP 调用端报错（RuntimeError），不降级，见三、降级行为表 |
+| ⚠️-2 | `run_decay_cycle()` auto_resolved 实现存疑 | ✅ 已确认实现 | `decay_engine.py` 完整实现 imp≤4 + >30天 + 未解决 → `bucket_mgr.update(resolved=True)` |
+| ⚠️-3 | `list_all()` 是否遍历 `feel/` 子目录 | ✅ 已确认实现 | `list_all()` dirs 明确包含 `self.feel_dir`，递归遍历 |
+| ⚠️-4 | `_time_ripple()` 浮点增量被 `int()` 截断 | ❌ 已确认 Bug | 见 B-03，决策见下 |
+| ⚠️-5 | Dashboard `/api/*` 路由认证覆盖 | ✅ 已确认覆盖 | 所有 `/api/buckets`、`/api/search`、`/api/network`、`/api/bucket/{id}`、`/api/breath-debug` 均调用 `_require_auth(request)` |
+
+---
+
+### 5.2 新发现 Bug 及修复决策
+
+| 编号 | 场景 | 严重度 | 问题描述 | 决策 & 修复方案 |
+|------|------|--------|----------|----------------|
+| **B-01** | 场景7a | 高 | `bucket_mgr.update(resolved=True)` 当前会将桶立即移入 `archive/`（type="archived"），规格预期"降权留存、关键词可激活"。resolved 桶实质上立即从所有搜索路径消失。 | **修复**：移除 `bucket_manager.py` `update()` 中 `resolved → _move_bucket(archive_dir)` 的自动归档逻辑，仅更新 frontmatter `resolved=True`，由 decay 引擎自然衰减至 archive。 |
+| **B-03** | 全局 | 高 | `_time_ripple()` 对 `activation_count` 做浮点增量（+0.3），但 `calculate_score()` 中 `max(1, int(...))` 截断小数，增量丢失，时间涟漪对衰减分无实际效果。 | **修复**：`decay_engine.py` `calculate_score()` 中改为 `activation_count = max(1.0, float(metadata.get("activation_count", 1)))` |
+| **B-04** | 场景1 | 中 | `bucket_manager.create()` 初始化 `activation_count=1`，冷启动检测条件 `activation_count==0` 对所有正常创建的桶永不满足，高重要度新桶不被优先浮现。 | **决策：初始化改为 `activation_count=0`**。语义上"创建"≠"被召回"，`touch()` 首次命中后变为 1，冷启动检测自然生效。规格已更新（见场景1步骤6 & 场景3 create 详情）。 |
+| **B-05** | 场景5 | 中 | `bucket_manager.py` `_calc_time_score()` 实现 `e^(-0.1×days)`，规格为 `e^(-0.02×days)`，衰减速度快 5 倍，30天后时间分 ≈ 0.05（规格预期 ≈ 0.55），旧记忆时间维度近乎失效。 | **决策：保留规格值 `0.02`**。记忆系统中旧记忆应通过关键词仍可被唤醒，时间维度是辅助信号不是淘汰信号。修复：`_calc_time_score()` 改为 `return math.exp(-0.02 * days)` |
+| **B-06** | 场景5 | 中 | `bucket_manager.py` `w_time` 默认值为 `2.5`，规格为 `1.5`，叠加 B-05 会导致时间维度严重偏重近期记忆。 | **决策：保留规格值 `1.5`**。修复：`w_time = scoring.get("time_proximity", 1.5)` |
+| **B-07** | 场景5 | 中 | `bucket_manager.py` `content_weight` 默认值为 `3.0`，规格为 `1.0`（body×1）。正文权重过高导致合并检测（`search(content, limit=1)`）误判——内容相似但主题不同的桶被错误合并。 | **决策：保留规格值 `1.0`**。正文是辅助信号，主要靠 name/tags/domain 识别同话题桶。修复：`content_weight = scoring.get("content_weight", 1.0)` |
+| **B-08** | 场景8 | 低 | `run_decay_cycle()` 内 auto_resolve 后继续使用旧 `meta` 变量计算 score，`resolved_factor=0.05` 需等下一 cycle 才生效。 | **修复**：auto_resolve 成功后执行 `meta["resolved"] = True` 刷新本地 meta 变量。 |
+| **B-09** | 场景3 | 低 | `hold()` 非 feel 路径中，用户显式传入的 `valence`/`arousal` 被 `analyze()` 返回值完全覆盖。 | **修复**：若用户显式传入（`0 <= valence <= 1`），优先使用用户值，`analyze()` 结果作为 fallback。 |
+| **B-10** | 场景10 | 低 | feel 桶以 `domain=[]` 创建，但 `bucket_manager.create()` 中 `domain or ["未分类"]` 兜底写入 `["未分类"]`，数据不干净。 | **修复**：`create()` 中对 `bucket_type=="feel"` 单独处理，允许空 domain 直接写入。 |
+
+---
+
+### 5.3 已确认正常实现
+
+- `breath()` 浮现模式不调用 `touch()`，不重置衰减计时器
+- `feel` 桶 `calculate_score()` 返回固定 50.0，永不归档
+- `breath(domain="feel")` 独立通道，按 `created` 降序，不压缩展示原文
+- `decay_engine.calculate_score()` 短期（≤3天）/ 长期（>3天）权重分离公式
+- `urgency_boost`：`arousal > 0.7 && !resolved → ×1.5`
+- `dream()` 连接提示（best_sim > 0.5）+ 结晶提示（feel 相似度 > 0.7 × ≥2 个）
+- 所有 `/api/*` Dashboard 路由均受 `_require_auth` 保护
+- `trace(delete=True)` 同步调用 `embedding_engine.delete_embedding()`
+- `grow()` 单条失败 `try/except` 隔离，标注 `⚠️条目名`，其他条继续
+
+---
+
+*本文档基于代码直接推导，每个步骤均可对照源文件函数名和行为验证。如代码更新，请同步修订此文档。*

ENV_VARS.md

@@ -0,0 +1,45 @@
+# 环境变量参考
+
+| 变量名 | 必填 | 默认值 | 说明 |
+|--------|------|--------|------|
+| `OMBRE_API_KEY` | 是 | — | Gemini / OpenAI-compatible API Key，用于脱水(dehydration)和向量嵌入 |
+| `OMBRE_BASE_URL` | 否 | `https://generativelanguage.googleapis.com/v1beta/openai/` | API Base URL（可替换为代理或兼容接口） |
+| `OMBRE_TRANSPORT` | 否 | `stdio` | MCP 传输模式：`stdio` / `sse` / `streamable-http` |
+| `OMBRE_PORT` | 否 | `8000` | HTTP/SSE 模式监听端口（仅 `sse` / `streamable-http` 生效） |
+| `OMBRE_BUCKETS_DIR` | 否 | `./buckets` | 记忆桶文件存放目录（绑定 Docker Volume 时务必设置） |
+| `OMBRE_HOOK_URL` | 否 | — | Breath/Dream Webhook 推送地址（POST JSON），留空则不推送 |
+| `OMBRE_HOOK_SKIP` | 否 | `false` | 设为 `true`/`1`/`yes` 跳过 Webhook 推送（即使 `OMBRE_HOOK_URL` 已设置） |
+| `OMBRE_DASHBOARD_PASSWORD` | 否 | — | 预设 Dashboard 访问密码；设置后覆盖文件存储的密码，首次访问不弹设置向导 |
+| `OMBRE_DEHYDRATION_MODEL` | 否 | `deepseek-chat` | 脱水/打标/合并/拆分用的 LLM 模型名（覆盖 `dehydration.model`） |
+| `OMBRE_DEHYDRATION_BASE_URL` | 否 | `https://api.deepseek.com/v1` | 脱水模型的 API Base URL（覆盖 `dehydration.base_url`） |
+| `OMBRE_MODEL` | 否 | — | `OMBRE_DEHYDRATION_MODEL` 的别名（前者优先） |
+| `OMBRE_EMBEDDING_MODEL` | 否 | `gemini-embedding-001` | 向量嵌入模型名（覆盖 `embedding.model`） |
+| `OMBRE_EMBEDDING_BASE_URL` | 否 | — | 向量嵌入的 API Base URL（覆盖 `embedding.base_url`；留空则复用脱水配置） |
+
+## 说明
+
+- `OMBRE_API_KEY` 也可在 `config.yaml` 的 `dehydration.api_key` / `embedding.api_key` 中设置，但**强烈建议**通过环境变量传入，避免密钥写入文件。
+- `OMBRE_DASHBOARD_PASSWORD` 设置后，Dashboard 的"修改密码"功能将被禁用（显示提示，建议直接修改环境变量）。未设置则密码存储在 `{buckets_dir}/.dashboard_auth.json`（SHA-256 + salt）。
+
+## Webhook 推送格式 (`OMBRE_HOOK_URL`)
+
+设置 `OMBRE_HOOK_URL` 后，Ombre Brain 会在以下事件发生时**异步**（fire-and-forget，5 秒超时）`POST` JSON 到该 URL：
+
+| 事件名 (`event`) | 触发时机 | `payload` 字段 |
+|------------------|----------|----------------|
+| `breath` | MCP 工具 `breath()` 返回时 | `mode` (`ok`/`empty`), `matches`, `chars` |
+| `dream` | MCP 工具 `dream()` 返回时 | `recent`, `chars` |
+| `breath_hook` | HTTP `GET /breath-hook` 命中（SessionStart 钩子） | `surfaced`, `chars` |
+| `dream_hook` | HTTP `GET /dream-hook` 命中 | `surfaced`, `chars` |
+
+请求体结构（JSON）：
+
+```json
+{
+  "event": "breath",
+  "timestamp": 1730000000.123,
+  "payload": { "...": "..." }
+}
+```
+
+Webhook 推送失败仅在服务日志中以 WARNING 级别记录，**不会影响 MCP 工具的正常返回**。

INTERNALS.md

@@ -65,7 +65,7 @@
 **自动化处理**
 - 存入时 LLM 自动分析 domain/valence/arousal/tags/name
 - 大段日记 LLM 拆分为 2~6 条独立记忆
-- 浮现时自动脱水压缩（LLM 压缩保语义，API 不可用降级到本地关键词提取）
+- 浮现时自动脱水压缩（LLM 压缩保语义，API 不可用时直接报错，无静默降级）
 - Wikilink `[[]]` 由 LLM 在内容中标记
 
 ---
@@ -76,7 +76,7 @@
 
 | 工具 | 关键参数 | 功能 |
 |---|---|---|
-| `breath` | query, max_tokens, domain, valence, arousal, max_results | 检索/浮现记忆 |
+| `breath` | query, max_tokens, domain, valence, arousal, max_results, **importance_min** | 检索/浮现记忆 |
 | `hold` | content, tags, importance, pinned, feel, source_bucket, valence, arousal | 存储记忆 |
 | `grow` | content | 日记拆分归档 |
 | `trace` | bucket_id, name, domain, valence, arousal, importance, tags, resolved, pinned, digested, content, delete | 修改元数据/内容/删除 |
@@ -85,10 +85,11 @@
 
 **工具详细行为**
 
-**`breath`** — 两种模式：
+**`breath`** — 三种模式：
-- **浮现模式**（无 query）：无参调用，按衰减引擎活跃度排序返回 top 记忆，permanent/pinned 始终浮现
+- **浮现模式**（无 query）：无参调用，按衰减引擎活跃度排序返回 top 记忆，钉选桶始终展示；冷启动检测（`activation_count==0 && importance>=8`）的桶最多 2 个插入最前，再 Top-1 固定 + Top-20 随机打乱
 - **检索模式**（有 query）：关键词 + 向量双通道搜索，四维评分（topic×4 + emotion×2 + time×2.5 + importance×1），阈值过滤
 - **Feel 检索**（`domain="feel"`）：特殊通道，按创建时间倒序返回所有 feel 类型桶，不走评分逻辑
+- **重要度批量模式**（`importance_min>=1`）：跳过语义搜索，直接筛选 importance≥importance_min 的桶，按 importance 降序，最多 20 条
 - 若指定 valence，对匹配桶的 valence 微调 ±0.1（情感记忆重构）
 
 **`hold`** — 两种模式：
@@ -120,26 +121,41 @@
 | `/breath-hook` | GET | SessionStart 钩子 |
 | `/dream-hook` | GET | Dream 钩子 |
 | `/dashboard` | GET | Dashboard 页面 |
-| `/api/buckets` | GET | 桶列表 |
+| `/api/buckets` | GET | 桶列表 🔒 |
-| `/api/bucket/{id}` | GET | 桶详情 |
+| `/api/bucket/{id}` | GET | 桶详情 🔒 |
-| `/api/search?q=` | GET | 搜索 |
+| `/api/search?q=` | GET | 搜索 🔒 |
-| `/api/network` | GET | 向量相似网络 |
+| `/api/network` | GET | 向量相似网络 🔒 |
-| `/api/breath-debug` | GET | 评分调试 |
+| `/api/breath-debug` | GET | 评分调试 🔒 |
-| `/api/config` | GET | 配置查看（key 脱敏） |
+| `/api/config` | GET | 配置查看（key 脱敏）🔒 |
-| `/api/config` | POST | 热更新配置 |
+| `/api/config` | POST | 热更新配置 🔒 |
-| `/api/import/upload` | POST | 上传并启动历史对话导入 |
+| `/api/status` | GET | 系统状态（版本/桶数/引擎）🔒 |
-| `/api/import/status` | GET | 导入进度查询 |
+| `/api/import/upload` | POST | 上传并启动历史对话导入 🔒 |
-| `/api/import/pause` | POST | 暂停/继续导入 |
+| `/api/import/status` | GET | 导入进度查询 🔒 |
-| `/api/import/patterns` | GET | 导入完成后词频规律检测 |
+| `/api/import/pause` | POST | 暂停/继续导入 🔒 |
-| `/api/import/results` | GET | 已导入记忆桶列表 |
+| `/api/import/patterns` | GET | 导入完成后词频规律检测 🔒 |
-| `/api/import/review` | POST | 批量审阅/批准导入结果 |
+| `/api/import/results` | GET | 已导入记忆桶列表 🔒 |
+| `/api/import/review` | POST | 批量审阅/批准导入结果 🔒 |
+| `/auth/status` | GET | 认证状态（是否需要初始化密码）|
+| `/auth/setup` | POST | 首次设置密码 |
+| `/auth/login` | POST | 密码登录，颁发 session cookie |
+| `/auth/logout` | POST | 注销 session |
+| `/auth/change-password` | POST | 修改密码 🔒 |
 
-**Dashboard（5 个 Tab）**
+> 🔒 = 需要 Dashboard 认证（未认证返回 401 JSON）
+
+**Dashboard 认证**
+- 密码存储：SHA-256 + 随机 salt，保存于 `{buckets_dir}/.dashboard_auth.json`
+- 环境变量 `OMBRE_DASHBOARD_PASSWORD` 设置后，覆盖文件密码（只读，不可通过 Dashboard 修改）
+- Session：内存字典（服务重启失效），cookie `ombre_session`（HttpOnly, SameSite=Lax, 7天）
+- `/health`, `/breath-hook`, `/dream-hook`, `/mcp*` 路径不受保护（公开）
+
+**Dashboard（6 个 Tab）**
 1. 记忆桶列表：6 种过滤器 + 主题域过滤 + 搜索 + 详情面板
 2. Breath 模拟：输入参数 → 可视化五步流程 → 四维条形图
 3. 记忆网络：Canvas 力导向图（节点=桶，边=相似度）
 4. 配置：热更新脱水/embedding/合并参数
 5. 导入：历史对话拖拽上传 → 分块处理进度条 → 词频规律分析 → 导入结果审阅
+6. 设置：服务状态监控、修改密码、退出登录
 
 **部署选项**
 1. 本地 stdio（`python server.py`）
@@ -152,7 +168,7 @@
 **迁移/批处理工具**：`migrate_to_domains.py`、`reclassify_domains.py`、`reclassify_api.py`、`backfill_embeddings.py`、`write_memory.py`、`check_buckets.py`、`import_memory.py`（历史对话导入引擎）
 
 **降级策略**
-- 脱水 API 不可用 → 本地关键词提取 + 句子评分
+- 脱水 API 不可用 → 直接抛 RuntimeError（设计决策，详见 BEHAVIOR_SPEC.md 三、降级行为表）
 - 向量搜索不可用 → 纯 fuzzy match
 - 逐条错误隔离（grow 中单条失败不影响其他）
 
@@ -172,6 +188,7 @@
 | `OMBRE_BUCKETS_DIR` | 记忆桶存储目录路径 | 否 | `""` → 回退到 config 或 `./buckets` |
 | `OMBRE_HOOK_URL` | SessionStart 钩子调用的服务器 URL | 否 | `"http://localhost:8000"` |
 | `OMBRE_HOOK_SKIP` | 设为 `"1"` 跳过 SessionStart 钩子 | 否 | 未设置（不跳过） |
+| `OMBRE_DASHBOARD_PASSWORD` | 预设 Dashboard 访问密码；设置后覆盖文件密码，首次访问不弹设置向导 | 否 | `""` |
 
 环境变量优先级：`环境变量 > config.yaml > 硬编码默认值`。所有环境变量在 `utils.py` 中读取并注入 config dict。
 
@@ -199,7 +216,7 @@
 | `server.py` | MCP 服务器主入口，注册工具 + Dashboard API + 钩子端点 | `bucket_manager`, `dehydrator`, `decay_engine`, `embedding_engine`, `utils` | `test_tools.py` |
 | `bucket_manager.py` | 记忆桶 CRUD、多维索引搜索、wikilink 注入、激活更新 | `utils` | `server.py`, `check_buckets.py`, `backfill_embeddings.py` |
 | `decay_engine.py` | 衰减引擎：遗忘曲线计算、自动归档、自动结案 | 无（接收 `bucket_mgr` 实例） | `server.py` |
-| `dehydrator.py` | 数据脱水压缩 + 合并 + 自动打标（LLM API + 本地降级） | `utils` | `server.py` |
+| `dehydrator.py` | 数据脱水压缩 + 合并 + 自动打标（仅 LLM API，不可用时报 RuntimeError） | `utils` | `server.py` |
 | `embedding_engine.py` | 向量化引擎：Gemini embedding API + SQLite + 余弦搜索 | `utils` | `server.py`, `backfill_embeddings.py` |
 | `utils.py` | 配置加载、日志、路径安全、ID 生成、token 估算 | 无 | 所有模块 |
 | `write_memory.py` | 手动写入记忆 CLI（绕过 MCP） | 无（独立脚本） | 无 |
@@ -372,12 +389,12 @@
 
 ### 5.4 为什么有 dehydration（脱水）这一层？
 
-**决策**：存入前先用 LLM 压缩内容（保留信息密度，去除冗余表达），API 不可用时降级到本地关键词提取。
+**决策**：存入前先用 LLM 压缩内容（保留信息密度，去除冗余表达）。API 不可用时直接抛出 `RuntimeError`，不静默降级。
 
 **理由**：
 - MCP 上下文有 token 限制，原始对话冗长，需要压缩
 - LLM 压缩能保留语义和情感色彩，纯截断会丢信息
-- 降级到本地确保离线可用——关键词提取 + 句子排序 + 截断
+- 本地关键词提取质量不足以替代语义打标与合并，静默降级会产生错误分类记忆，比报错更危险。详见 BEHAVIOR_SPEC.md 三、降级行为表。
 
 **放弃方案**：只做截断。信息损失太大。
 
@@ -479,3 +496,92 @@ type: dynamic
 
 桶正文内容...
 ```
+
+---
+
+## 7. Bug 修复记录 (B-01 至 B-10)
+
+### B-01 — `update(resolved=True)` 自动归档 🔴 高
+
+- **文件**: `bucket_manager.py` → `update()`
+- **问题**: `resolved=True` 时立即调用 `_move_bucket(archive_dir)` 将桶移入 `archive/`
+- **修复**: 移除 `_move_bucket` 逻辑；resolved 桶留在 `dynamic/`，由 decay 引擎自然淘汰
+- **影响**: 已解决的桶仍可被关键词检索命中（降权但不消失）
+- **测试**: `tests/regression/test_issue_B01.py`，`tests/integration/test_scenario_07_trace.py`
+
+### B-03 — `int()` 截断浮点 activation_count 🔴 高
+
+- **文件**: `decay_engine.py` → `calculate_score()`
+- **问题**: `max(1, int(activation_count))` 将 `_time_ripple` 写入的 1.3 截断为 1，涟漪加成失效
+- **修复**: 改为 `max(1.0, float(activation_count))`
+- **影响**: 时间涟漪效果现在正确反映在 score 上；高频访问的桶衰减更慢
+- **测试**: `tests/regression/test_issue_B03.py`，`tests/unit/test_calculate_score.py`
+
+### B-04 — `create()` 初始化 activation_count=1 🟠 中
+
+- **文件**: `bucket_manager.py` → `create()`
+- **问题**: `activation_count=1` 导致冷启动检测条件 `== 0` 永不满足，新建重要桶无法浮现
+- **修复**: 改为 `activation_count=0`；`touch()` 首次命中后变 1
+- **测试**: `tests/regression/test_issue_B04.py`，`tests/integration/test_scenario_01_cold_start.py`
+
+### B-05 — 时间衰减系数 0.1 过快 🟠 中
+
+- **文件**: `bucket_manager.py` → `_calc_time_score()`
+- **问题**: `math.exp(-0.1 * days)` 导致 30 天后得分仅剩 ≈0.05，远快于人类记忆曲线
+- **修复**: 改为 `math.exp(-0.02 * days)`（30 天后 ≈0.549）
+- **影响**: 记忆保留时间更符合人类认知模型
+- **测试**: `tests/regression/test_issue_B05.py`，`tests/unit/test_score_components.py`
+
+### B-06 — `w_time` 默认值 2.5 过高 🟠 中
+
+- **文件**: `bucket_manager.py` → `_calc_final_score()`（或评分调用处）
+- **问题**: `scoring.get("time_proximity", 2.5)` — 时间权重过高，近期低质量记忆得分高于高质量旧记忆
+- **修复**: 改为 `scoring.get("time_proximity", 1.5)`
+- **测试**: `tests/regression/test_issue_B06.py`，`tests/unit/test_score_components.py`
+
+### B-07 — `content_weight` 默认值 3.0 过高 🟠 中
+
+- **文件**: `bucket_manager.py` → `_calc_topic_score()`
+- **问题**: `scoring.get("content_weight", 3.0)` — 内容权重远大于名字权重(×3)，导致内容重复堆砌的桶得分高于名字精确匹配的桶
+- **修复**: 改为 `scoring.get("content_weight", 1.0)`
+- **影响**: 名字完全匹配 > 标签匹配 > 内容匹配的得分层级现在正确
+- **测试**: `tests/regression/test_issue_B07.py`，`tests/unit/test_topic_score.py`
+
+### B-08 — `run_decay_cycle()` 同轮 auto_resolve 后 score 未降权 🟡 低
+
+- **文件**: `decay_engine.py` → `run_decay_cycle()`
+- **问题**: `auto_resolve` 标记后立即用旧 `meta`（stale）计算 score，`resolved_factor=0.05` 未生效
+- **修复**: 在 `bucket_mgr.update(resolved=True)` 后立即执行 `meta["resolved"] = True`，确保同轮降权
+- **测试**: `tests/regression/test_issue_B08.py`，`tests/integration/test_scenario_08_decay.py`
+
+### B-09 — `hold()` 用 analyze() 覆盖用户传入的 valence/arousal 🟡 低
+
+- **文件**: `server.py` → `hold()`
+- **问题**: 先调 `analyze()`，再直接用结果覆盖用户传入的情感值，情感准确性丢失
+- **修复**: 使用 `final_valence = user_valence if user_valence is not None else analyze_result.get("valence")`
+- **影响**: 用户明确传入的情感坐标（包括 0.0）不再被 LLM 结果覆盖
+- **测试**: `tests/regression/test_issue_B09.py`，`tests/integration/test_scenario_03_hold.py`
+
+### B-10 — feel 桶 `domain=[]` 被填充为 `["未分类"]` 🟡 低
+
+- **文件**: `bucket_manager.py` → `create()`
+- **问题**: `if not domain: domain = ["未分类"]` 对所有桶类型生效，feel 桶的空 domain 被错误填充
+- **修复**: 改为 `if not domain and bucket_type != "feel": domain = ["未分类"]`
+- **影响**: `breath(domain="feel")` 通道过滤逻辑现在正确（feel 桶 domain 始终为空列表）
+- **测试**: `tests/regression/test_issue_B10.py`，`tests/integration/test_scenario_10_feel.py`
+
+---
+
+### Bug 修复汇总表
+
+| ID | 严重度 | 文件 | 方法 | 一句话描述 |
+|---|---|---|---|---|
+| B-01 | 🔴 高 | `bucket_manager.py` | `update()` | resolved 桶不再自动归档 |
+| B-03 | 🔴 高 | `decay_engine.py` | `calculate_score()` | float activation_count 不被 int() 截断 |
+| B-04 | 🟠 中 | `bucket_manager.py` | `create()` | 初始 activation_count=0 |
+| B-05 | 🟠 中 | `bucket_manager.py` | `_calc_time_score()` | 时间衰减系数 0.02（原 0.1） |
+| B-06 | 🟠 中 | `bucket_manager.py` | 评分权重配置 | w_time 默认 1.5（原 2.5） |
+| B-07 | 🟠 中 | `bucket_manager.py` | `_calc_topic_score()` | content_weight 默认 1.0（原 3.0） |
+| B-08 | 🟡 低 | `decay_engine.py` | `run_decay_cycle()` | auto_resolve 同轮应用 ×0.05 |
+| B-09 | 🟡 低 | `server.py` | `hold()` | 用户 valence/arousal 优先 |
+| B-10 | 🟡 低 | `bucket_manager.py` | `create()` | feel 桶 domain=[] 不被填充 |

README.md

@@ -159,7 +159,7 @@ OMBRE_API_KEY=你的API密钥
 > 3. Set `dehydration.base_url` to `https://generativelanguage.googleapis.com/v1beta/openai` in `config.yaml`
 > Also supports DeepSeek, Ollama, LM Studio, vLLM, or any OpenAI-compatible API.
 
-没有 API key 也能用，脱水压缩会降级到本地模式，只是效果差一点。那就写：
+没有 API key 则脱水压缩和自动打标功能不可用（会报错），但记忆的读写和检索仍正常工作。如果暂时不用脱水功能，可以留空：
 
 ```
 OMBRE_API_KEY=
@@ -435,6 +435,27 @@ Sensitive config via env vars:
 - `OMBRE_API_KEY` — LLM API 密钥
 - `OMBRE_TRANSPORT` — 覆盖传输方式
 - `OMBRE_BUCKETS_DIR` — 覆盖存储路径
+- `OMBRE_DASHBOARD_PASSWORD` — Dashboard 访问密码（可选，见下）
+
+## Dashboard 认证 / Dashboard Auth
+
+自 v1.3.0 起，Dashboard 和所有 `/api/*` 端点均受密码保护。
+Since v1.3.0, the Dashboard and all `/api/*` endpoints are password-protected.
+
+**首次访问**：若未设置密码，浏览器会弹出设置向导，填写并确认密码后即可使用。
+**First visit**: If no password is set, a setup wizard will appear. Enter and confirm a password to get started.
+
+**通过环境变量预设密码**：在 `docker-compose.user.yml` 中添加：
+**Pre-set via env var** in your `docker-compose.user.yml`:
+```yaml
+environment:
+  - OMBRE_DASHBOARD_PASSWORD=your_password_here
+```
+设置后，Dashboard 的"修改密码"功能将被禁用，必须通过环境变量修改。
+When set, the in-Dashboard password change is disabled — modify the env var directly.
+
+完整环境变量说明见 [ENV_VARS.md](ENV_VARS.md)。
+Full env var reference: [ENV_VARS.md](ENV_VARS.md).
 
 ## 衰减公式 / Decay Formula
 
@@ -570,14 +591,14 @@ Dashboard：浏览器打开 `http://localhost:8000/dashboard`
 > **Free tier won't work**: Render free tier has **no persistent disk** — all memory data is lost on restart. It also sleeps on inactivity. **Starter plan ($7/mo) or above is required.**
 
 项目根目录已包含 `render.yaml`，点击按钮后：
-1. （可选）设置 `OMBRE_API_KEY`：任何 OpenAI 兼容 API 的 key，不填则自动降级为本地关键词提取
+1. 设置 `OMBRE_API_KEY`：任何 OpenAI 兼容 API 的 key（**必需**，未设置时 hold/grow 会报错、仅检索类工具可用）
 2. （可选）设置 `OMBRE_BASE_URL`：API 地址，支持任意 OpenAI 化地址，如 `https://api.deepseek.com/v1` / `http://123.1.1.1:7689/v1` / `http://your-ollama:11434/v1`
 3. Render 自动挂载持久化磁盘到 `/opt/render/project/src/buckets`
 4. Dashboard：`https://<你的服务名>.onrender.com/dashboard`
 5. 部署后 MCP URL：`https://<你的服务名>.onrender.com/mcp`
 
 `render.yaml` is included. After clicking the button:
-1. (Optional) `OMBRE_API_KEY`: any OpenAI-compatible key; omit to fall back to local keyword extraction
+1. `OMBRE_API_KEY`: any OpenAI-compatible key (**required** for hold/grow; without it those tools raise an error)
 2. (Optional) `OMBRE_BASE_URL`: any OpenAI-compatible endpoint, e.g. `https://api.deepseek.com/v1`, `http://123.1.1.1:7689/v1`, `http://your-ollama:11434/v1`
 3. Persistent disk auto-mounts at `/opt/render/project/src/buckets`
 4. Dashboard: `https://<your-service>.onrender.com/dashboard`
@@ -599,7 +620,7 @@ Dashboard：浏览器打开 `http://localhost:8000/dashboard`
    - Zeabur auto-detects the `Dockerfile` in root and builds via Docker
 
 2. **设置环境变量 / Set environment variables**（服务页面 → **Variables** 标签页）
-   - `OMBRE_API_KEY`（可选）— LLM API 密钥，不填则自动降级为本地关键词提取
+   - `OMBRE_API_KEY`（**必需**）— LLM API 密钥；未设置时 hold/grow/dream 会报错
    - `OMBRE_BASE_URL`（可选）— API 地址，如 `https://api.deepseek.com/v1`
 
    > ⚠️ **不需要**手动设置 `OMBRE_TRANSPORT` 和 `OMBRE_BUCKETS_DIR`，Dockerfile 里已经设好了默认值。Zeabur 对单阶段 Dockerfile 会自动注入控制台设置的环境变量。
@@ -783,6 +804,51 @@ sudo systemctl restart ombre-brain   # 示例
 > - If `requirements.txt` changed, Docker rebuild handles it automatically; non-Docker users need `pip install -r requirements.txt`
 > - After updating, visit `/health` to verify the service is running
 
+## 测试 / Testing
+
+测试套件覆盖规格书所有场景（场景 01–11），以及 B-01 至 B-10 全部 bug 修复的回归测试。
+
+The test suite covers all spec scenarios (01–11) and regression tests for every bug fix (B-01 to B-10).
+
+### 快速运行 / Quick Start
+
+```bash
+pip install pytest pytest-asyncio
+pytest tests/                          # 全部测试
+pytest tests/unit/                     # 单元测试
+pytest tests/integration/             # 集成测试（场景全流程）
+pytest tests/regression/              # 回归测试（B-01..B-10）
+pytest tests/ -k "B01"               # 单个回归测试
+pytest tests/ -v                       # 详细输出
+```
+
+### 测试层级 / Test Layers
+
+| 目录 Directory | 内容 Contents |
+|---|---|
+| `tests/unit/` | 单独测试 calculate_score、topic_score、时间得分、CRUD 等核心函数 |
+| `tests/integration/` | 场景全流程：冷启动、hold、search、trace、decay、feel 等 11 个场景 |
+| `tests/regression/` | 每个 bug（B-01 至 B-10）独立回归测试，含边界条件 |
+
+### 回归测试覆盖 / Regression Coverage
+
+| 文件 | Bug | 核心断言 |
+|---|---|---|
+| `test_issue_B01.py` | resolved 桶不再自动归档 | `update(resolved=True)` 后桶留在 `dynamic/`，搜索仍可命中，得分 ×0.05 |
+| `test_issue_B03.py` | float activation_count 不被 int() 截断 | 1.3 > 1.0 得分，`_time_ripple` 写入 0.3 增量 |
+| `test_issue_B04.py` | create() 初始 activation_count=0 | 新建桶满足冷启动条件，touch() 后变 1 |
+| `test_issue_B05.py` | 时间衰减系数 0.02（原 0.1）| 30天 ≈ 0.549，非旧值 0.049 |
+| `test_issue_B06.py` | w_time 默认 1.5（原 2.5）| `BucketManager.w_time == 1.5` |
+| `test_issue_B07.py` | content_weight 默认 1.0（原 3.0）| 名字完全匹配得分 > 内容模糊匹配 |
+| `test_issue_B08.py` | auto_resolve 同轮应用降权因子 | stale meta 修复后 score ×0.05 立即生效 |
+| `test_issue_B09.py` | hold() 保留用户传入的 valence/arousal | 用户值优先于 analyze() 结果 |
+| `test_issue_B10.py` | feel 桶 domain=[] 不被填充 | feel 桶保持 `[]`；dynamic 桶正确填 `["未分类"]` |
+
+> **测试隔离**：所有测试运行在 `tmp_path` 临时目录，绝不触碰真实记忆数据。
+> **Test isolation**: All tests run in `tmp_path` — your real memory data is never touched.
+
+---
+
 ## License
 
 MIT

backfill_embeddings.py

@@ -13,7 +13,6 @@ Free tier: 1500 requests/day, so ~75 batches of 20.
 import asyncio
 import argparse
 import sys
-import time
 
 sys.path.insert(0, ".")
 from utils import load_config
@@ -79,7 +78,7 @@ async def backfill(batch_size: int = 20, dry_run: bool = False):
                 print(f"  ERROR: {b['id'][:12]} ({name[:30]}): {e}")
 
         if i + batch_size < total:
-            print(f"  Waiting 2s before next batch...")
+            print("  Waiting 2s before next batch...")
             await asyncio.sleep(2)
 
     print(f"\n=== Done: {success} success, {failed} failed, {total - success - failed} skipped ===")

backup_20260405_2124/README.md

@@ -1,205 +0,0 @@
-# Ombre Brain
-
-一个给 Claude 用的长期情绪记忆系统。基于 Russell 效价/唤醒度坐标打标，Obsidian 做存储层，MCP 接入，带遗忘曲线。
-
-A long-term emotional memory system for Claude. Tags memories using Russell's valence/arousal coordinates, stores them as Obsidian-compatible Markdown, connects via MCP, and has a forgetting curve.
-
----
-
-## 它是什么 / What is this
-
-Claude 没有跨对话记忆。每次对话结束，之前聊过的所有东西都会消失。
-
-Ombre Brain 给了它一套持久记忆——不是那种冷冰冰的键值存储，而是带情感坐标的、会自然衰减的、像人类记忆一样会遗忘和浮现的系统。
-
-Claude has no cross-conversation memory. Everything from a previous chat vanishes once it ends.
-
-Ombre Brain gives it persistent memory — not cold key-value storage, but a system with emotional coordinates, natural decay, and forgetting/surfacing mechanics that loosely mimic how human memory works.
-
-核心特点 / Key features:
-
-- **情感坐标打标 / Emotional tagging**: 每条记忆用 Russell 环形情感模型的 valence（效价）和 arousal（唤醒度）两个连续维度标记。不是"开心/难过"这种离散标签。
-  Each memory is tagged with two continuous dimensions from Russell's circumplex model: valence and arousal. Not discrete labels like "happy/sad".
-
-- **自然遗忘 / Natural forgetting**: 改进版艾宾浩斯遗忘曲线。不活跃的记忆自动衰减归档，高情绪强度的记忆衰减更慢。
-  Modified Ebbinghaus forgetting curve. Inactive memories naturally decay and archive. High-arousal memories decay slower.
-
-- **权重池浮现 / Weight pool surfacing**: 记忆不是被动检索的，它们会主动浮现——未解决的、情绪强烈的记忆权重更高，会在对话开头自动推送。
-  Memories aren't just passively retrieved — they actively surface. Unresolved, emotionally intense memories carry higher weight and get pushed at conversation start.
-
-- **Obsidian 原生 / Obsidian-native**: 每个记忆桶就是一个 Markdown 文件，YAML frontmatter 存元数据。可以直接在 Obsidian 里浏览、编辑、搜索。自动注入 `[[双链]]`。
-  Each memory bucket is a Markdown file with YAML frontmatter. Browse, edit, and search directly in Obsidian. Wikilinks are auto-injected.
-
-- **API 降级 / API degradation**: 脱水压缩和自动打标优先用廉价 LLM API（DeepSeek 等），API 不可用时自动降级到本地关键词分析——始终可用。
-  Dehydration and auto-tagging prefer a cheap LLM API (DeepSeek etc.). When the API is unavailable, it degrades to local keyword analysis — always functional.
-
-## 边界说明 / Design boundaries
-
-官方记忆功能已经在做身份层的事了——你是谁，你有什么偏好，你们的关系是什么。那一层交给它，Ombre Brain不打算造重复的轮子。
-
-Ombre Brain 的边界是时间里发生的事，不是你是谁。它记住的是：你们聊过什么，经历了什么，哪些事情还悬在那里没有解决。两层配合用，才是完整的。
-
-每次新对话，Claude 从零开始——但它能从 Ombre Brain 里找回跟你有关的一切。不是重建，是接续。
-
----
-
-Official memory already handles the identity layer — who you are, what you prefer, what your relationship is. That layer belongs there. Ombre Brain isn't trying to duplicate it.
-
-Ombre Brain's boundary is *what happened in time*, not *who you are*. It holds conversations, experiences, unresolved things. The two layers together are what make it feel complete.
-
-Each new conversation starts fresh — but Claude can reach back through Ombre Brain and find everything that happened between you. Not a rebuild. A continuation.
-
-## 架构 / Architecture
-
-```
-Claude ←→ MCP Protocol ←→ server.py
-                              │
-              ┌───────────────┼───────────────┐
-              │               │               │
-        bucket_manager   dehydrator     decay_engine
-         (CRUD + 搜索)    (压缩 + 打标)   (遗忘曲线)
-              │
-        Obsidian Vault (Markdown files)
-```
-
-5 个 MCP 工具 / 5 MCP tools:
-
-| 工具 Tool | 作用 Purpose |
-|-----------|-------------|
-| `breath` | 浮现或检索记忆。无参数=推送未解决记忆；有参数=关键词+情感检索 / Surface or search memories |
-| `hold` | 存储单条记忆，自动打标+合并相似桶 / Store a single memory with auto-tagging |
-| `grow` | 日记归档，自动拆分长内容为多个记忆桶 / Diary digest, auto-split into multiple buckets |
-| `trace` | 修改元数据、标记已解决、删除 / Modify metadata, mark resolved, delete |
-| `pulse` | 系统状态 + 所有记忆桶列表 / System status + bucket listing |
-
-## 安装 / Setup
-
-### 环境要求 / Requirements
-
-- Python 3.11+
-- 一个 Obsidian Vault（可选，不用也行，会在项目目录下自建 `buckets/`）
-  An Obsidian vault (optional — without one, it uses a local `buckets/` directory)
-
-### 步骤 / Steps
-
-```bash
-git clone https://github.com/P0lar1zzZ/Ombre-Brain.git
-cd Ombre-Brain
-
-python -m venv .venv
-source .venv/bin/activate  # Windows: .venv\Scripts\activate
-
-pip install -r requirements.txt
-```
-
-复制配置文件并按需修改 / Copy config and edit as needed:
-
-```bash
-cp config.example.yaml config.yaml
-```
-
-如果你要用 API 做脱水压缩和自动打标（推荐，效果好很多），设置环境变量：
-If you want API-powered dehydration and tagging (recommended, much better quality):
-
-```bash
-export OMBRE_API_KEY="your-api-key"
-```
-
-支持任何 OpenAI 兼容 API。在 `config.yaml` 里改 `base_url` 和 `model` 就行。
-Supports any OpenAI-compatible API. Just change `base_url` and `model` in `config.yaml`.
-
-### 接入 Claude Desktop / Connect to Claude Desktop
-
-在 Claude Desktop 配置文件中添加（macOS: `~/Library/Application Support/Claude/claude_desktop_config.json`）：
-
-Add to your Claude Desktop config:
-
-```json
-{
-  "mcpServers": {
-    "ombre-brain": {
-      "command": "python",
-      "args": ["/path/to/Ombre-Brain/server.py"],
-      "env": {
-        "OMBRE_API_KEY": "your-api-key"
-      }
-    }
-  }
-}
-```
-
-### 接入 Claude.ai (远程) / Connect to Claude.ai (remote)
-
-需要 HTTP 传输 + 隧道。可以用 Docker：
-Requires HTTP transport + tunnel. Docker setup:
-
-```bash
-echo "OMBRE_API_KEY=your-api-key" > .env
-docker-compose up -d
-```
-
-`docker-compose.yml` 里配好了 Cloudflare Tunnel。你需要自己在 `~/.cloudflared/` 下放凭证和路由配置。
-The `docker-compose.yml` includes Cloudflare Tunnel. You'll need your own credentials under `~/.cloudflared/`.
-
-### 指向 Obsidian / Point to Obsidian
-
-在 `config.yaml` 里设置 `buckets_dir`：
-Set `buckets_dir` in `config.yaml`:
-
-```yaml
-buckets_dir: "/path/to/your/Obsidian Vault/Ombre Brain"
-```
-
-不设的话，默认用项目目录下的 `buckets/`。
-If not set, defaults to `buckets/` in the project directory.
-
-## 配置 / Configuration
-
-所有参数在 `config.yaml`（从 `config.example.yaml` 复制）。关键的几个：
-All parameters in `config.yaml` (copy from `config.example.yaml`). Key ones:
-
-| 参数 Parameter | 说明 Description | 默认 Default |
-|---|---|---|
-| `transport` | `stdio`（本地）/ `streamable-http`（远程）| `stdio` |
-| `buckets_dir` | 记忆桶存储路径 / Bucket storage path | `./buckets/` |
-| `dehydration.model` | 脱水用的 LLM 模型 / LLM model for dehydration | `deepseek-chat` |
-| `dehydration.base_url` | API 地址 / API endpoint | `https://api.deepseek.com/v1` |
-| `decay.lambda` | 衰减速率，越大越快忘 / Decay rate | `0.05` |
-| `decay.threshold` | 归档阈值 / Archive threshold | `0.3` |
-| `merge_threshold` | 合并相似度阈值 (0-100) / Merge similarity | `75` |
-
-敏感配置用环境变量：
-Sensitive config via env vars:
-- `OMBRE_API_KEY` — LLM API 密钥
-- `OMBRE_TRANSPORT` — 覆盖传输方式
-- `OMBRE_BUCKETS_DIR` — 覆盖存储路径
-
-## 衰减公式 / Decay Formula
-
-$$Score = Importance \times activation\_count^{0.3} \times e^{-\lambda \times days} \times (base + arousal \times boost)$$
-
-- `importance`: 1-10，记忆重要性 / memory importance
-- `activation_count`: 被检索的次数，越常被想起衰减越慢 / retrieval count; more recalls = slower decay
-- `days`: 距上次激活的天数 / days since last activation
-- `arousal`: 唤醒度，越强烈的记忆越难忘 / arousal; intense memories are harder to forget
-- 已解决的记忆权重降到 5%，沉底等被关键词唤醒 / resolved memories drop to 5%, sink until keyword-triggered
-
-## 给 Claude 的使用指南 / Usage Guide for Claude
-
-`CLAUDE_PROMPT.md` 是写给 Claude 看的使用说明。放到你的 system prompt 或 custom instructions 里就行。
-
-`CLAUDE_PROMPT.md` is the usage guide written for Claude. Put it in your system prompt or custom instructions.
-
-## 工具脚本 / Utility Scripts
-
-| 脚本 Script | 用途 Purpose |
-|---|---|
-| `write_memory.py` | 手动写入记忆，绕过 MCP / Manually write memories, bypass MCP |
-| `migrate_to_domains.py` | 迁移平铺文件到域子目录 / Migrate flat files to domain subdirs |
-| `reclassify_domains.py` | 基于关键词重分类 / Reclassify by keywords |
-| `reclassify_api.py` | 用 API 重打标未分类桶 / Re-tag uncategorized buckets via API |
-| `test_smoke.py` | 冒烟测试 / Smoke test |
-
-## License
-
-MIT

backup_20260405_2124/bucket_manager.py

@@ -1,755 +0,0 @@
-# ============================================================
-# Module: Memory Bucket Manager (bucket_manager.py)
-# 模块：记忆桶管理器
-#
-# CRUD operations, multi-dimensional index search, activation updates
-# for memory buckets.
-# 记忆桶的增删改查、多维索引搜索、激活更新。
-#
-# Core design:
-# 核心逻辑：
-#   - Each bucket = one Markdown file (YAML frontmatter + body)
-#     每个记忆桶 = 一个 Markdown 文件
-#   - Storage by type: permanent / dynamic / archive
-#     存储按类型分目录
-#   - Multi-dimensional soft index: domain + valence/arousal + fuzzy text
-#     多维软索引：主题域 + 情感坐标 + 文本模糊匹配
-#   - Search strategy: domain pre-filter → weighted multi-dim ranking
-#     搜索策略：主题域预筛 → 多维加权精排
-#   - Emotion coordinates based on Russell circumplex model:
-#     情感坐标基于环形情感模型（Russell circumplex）：
-#       valence (0~1): 0=negative → 1=positive
-#       arousal (0~1): 0=calm → 1=excited
-#
-# Depended on by: server.py, decay_engine.py
-# 被谁依赖：server.py, decay_engine.py
-# ============================================================
-
-import os
-import math
-import logging
-import re
-import shutil
-from collections import Counter
-from datetime import datetime
-from pathlib import Path
-from typing import Optional
-
-import frontmatter
-import jieba
-from rapidfuzz import fuzz
-
-from utils import generate_bucket_id, sanitize_name, safe_path, now_iso
-
-logger = logging.getLogger("ombre_brain.bucket")
-
-
-class BucketManager:
-    """
-    Memory bucket manager — entry point for all bucket CRUD operations.
-    Buckets are stored as Markdown files with YAML frontmatter for metadata
-    and body for content. Natively compatible with Obsidian browsing/editing.
-    记忆桶管理器 —— 所有桶的 CRUD 操作入口。
-    桶以 Markdown 文件存储，YAML frontmatter 存元数据，正文存内容。
-    天然兼容 Obsidian 直接浏览和编辑。
-    """
-
-    def __init__(self, config: dict):
-        # --- Read storage paths from config / 从配置中读取存储路径 ---
-        self.base_dir = config["buckets_dir"]
-        self.permanent_dir = os.path.join(self.base_dir, "permanent")
-        self.dynamic_dir = os.path.join(self.base_dir, "dynamic")
-        self.archive_dir = os.path.join(self.base_dir, "archive")
-        self.fuzzy_threshold = config.get("matching", {}).get("fuzzy_threshold", 50)
-        self.max_results = config.get("matching", {}).get("max_results", 5)
-
-        # --- Wikilink config / 双链配置 ---
-        wikilink_cfg = config.get("wikilink", {})
-        self.wikilink_enabled = wikilink_cfg.get("enabled", True)
-        self.wikilink_use_tags = wikilink_cfg.get("use_tags", False)
-        self.wikilink_use_domain = wikilink_cfg.get("use_domain", True)
-        self.wikilink_use_auto_keywords = wikilink_cfg.get("use_auto_keywords", True)
-        self.wikilink_auto_top_k = wikilink_cfg.get("auto_top_k", 8)
-        self.wikilink_min_len = wikilink_cfg.get("min_keyword_len", 2)
-        self.wikilink_exclude_keywords = set(wikilink_cfg.get("exclude_keywords", []))
-        self.wikilink_stopwords = {
-            "的", "了", "在", "是", "我", "有", "和", "就", "不", "人",
-            "都", "一个", "上", "也", "很", "到", "说", "要", "去",
-            "你", "会", "着", "没有", "看", "好", "自己", "这", "他", "她",
-            "我们", "你们", "他们", "然后", "今天", "昨天", "明天", "一下",
-            "the", "and", "for", "are", "but", "not", "you", "all", "can",
-            "had", "her", "was", "one", "our", "out", "has", "have", "with",
-            "this", "that", "from", "they", "been", "said", "will", "each",
-        }
-        self.wikilink_stopwords |= {w.lower() for w in self.wikilink_exclude_keywords}
-
-        # --- Search scoring weights / 检索权重配置 ---
-        scoring = config.get("scoring_weights", {})
-        self.w_topic = scoring.get("topic_relevance", 4.0)
-        self.w_emotion = scoring.get("emotion_resonance", 2.0)
-        self.w_time = scoring.get("time_proximity", 1.5)
-        self.w_importance = scoring.get("importance", 1.0)
-
-    # ---------------------------------------------------------
-    # Create a new bucket
-    # 创建新桶
-    # Write content and metadata into a .md file
-    # 将内容和元数据写入一个 .md 文件
-    # ---------------------------------------------------------
-    async def create(
-        self,
-        content: str,
-        tags: list[str] = None,
-        importance: int = 5,
-        domain: list[str] = None,
-        valence: float = 0.5,
-        arousal: float = 0.3,
-        bucket_type: str = "dynamic",
-        name: str = None,
-    ) -> str:
-        """
-        Create a new memory bucket, return bucket ID.
-        创建一个新的记忆桶，返回桶 ID。
-        """
-        bucket_id = generate_bucket_id()
-        bucket_name = sanitize_name(name) if name else bucket_id
-        domain = domain or ["未分类"]
-        tags = tags or []
-        linked_content = self._apply_wikilinks(content, tags, domain, bucket_name)
-
-        # --- Build YAML frontmatter metadata / 构建元数据 ---
-        metadata = {
-            "id": bucket_id,
-            "name": bucket_name,
-            "tags": tags,
-            "domain": domain,
-            "valence": max(0.0, min(1.0, valence)),
-            "arousal": max(0.0, min(1.0, arousal)),
-            "importance": max(1, min(10, importance)),
-            "type": bucket_type,
-            "created": now_iso(),
-            "last_active": now_iso(),
-            "activation_count": 1,
-        }
-
-        # --- Assemble Markdown file (frontmatter + body) ---
-        # --- 组装 Markdown 文件 ---
-        post = frontmatter.Post(linked_content, **metadata)
-
-        # --- Choose directory by type + primary domain ---
-        # --- 按类型 + 主题域选择存储目录 ---
-        type_dir = self.permanent_dir if bucket_type == "permanent" else self.dynamic_dir
-        primary_domain = sanitize_name(domain[0]) if domain else "未分类"
-        target_dir = os.path.join(type_dir, primary_domain)
-        os.makedirs(target_dir, exist_ok=True)
-
-        # --- Filename: readable_name_bucketID.md (Obsidian friendly) ---
-        # --- 文件名：可读名称_桶ID.md ---
-        if bucket_name and bucket_name != bucket_id:
-            filename = f"{bucket_name}_{bucket_id}.md"
-        else:
-            filename = f"{bucket_id}.md"
-        file_path = safe_path(target_dir, filename)
-
-        try:
-            with open(file_path, "w", encoding="utf-8") as f:
-                f.write(frontmatter.dumps(post))
-        except OSError as e:
-            logger.error(f"Failed to write bucket file / 写入桶文件失败: {file_path}: {e}")
-            raise
-
-        logger.info(
-            f"Created bucket / 创建记忆桶: {bucket_id} ({bucket_name}) → {primary_domain}/"
-        )
-        return bucket_id
-
-    # ---------------------------------------------------------
-    # Read bucket content
-    # 读取桶内容
-    # Returns {"id", "metadata", "content", "path"} or None
-    # ---------------------------------------------------------
-    async def get(self, bucket_id: str) -> Optional[dict]:
-        """
-        Read a single bucket by ID.
-        根据 ID 读取单个桶。
-        """
-        if not bucket_id or not isinstance(bucket_id, str):
-            return None
-        file_path = self._find_bucket_file(bucket_id)
-        if not file_path:
-            return None
-        return self._load_bucket(file_path)
-
-    # ---------------------------------------------------------
-    # Update bucket
-    # 更新桶
-    # Supports: content, tags, importance, valence, arousal, name, resolved
-    # ---------------------------------------------------------
-    async def update(self, bucket_id: str, **kwargs) -> bool:
-        """
-        Update bucket content or metadata fields.
-        更新桶的内容或元数据字段。
-        """
-        file_path = self._find_bucket_file(bucket_id)
-        if not file_path:
-            return False
-
-        try:
-            post = frontmatter.load(file_path)
-        except Exception as e:
-            logger.warning(f"Failed to load bucket for update / 加载桶失败: {file_path}: {e}")
-            return False
-
-        # --- Update only fields that were passed in / 只改传入的字段 ---
-        if "content" in kwargs:
-            next_tags = kwargs.get("tags", post.get("tags", []))
-            next_domain = kwargs.get("domain", post.get("domain", []))
-            next_name = kwargs.get("name", post.get("name", ""))
-            post.content = self._apply_wikilinks(
-                kwargs["content"],
-                next_tags,
-                next_domain,
-                next_name,
-            )
-        if "tags" in kwargs:
-            post["tags"] = kwargs["tags"]
-        if "importance" in kwargs:
-            post["importance"] = max(1, min(10, int(kwargs["importance"])))
-        if "domain" in kwargs:
-            post["domain"] = kwargs["domain"]
-        if "valence" in kwargs:
-            post["valence"] = max(0.0, min(1.0, float(kwargs["valence"])))
-        if "arousal" in kwargs:
-            post["arousal"] = max(0.0, min(1.0, float(kwargs["arousal"])))
-        if "name" in kwargs:
-            post["name"] = sanitize_name(kwargs["name"])
-        if "resolved" in kwargs:
-            post["resolved"] = bool(kwargs["resolved"])
-
-        # --- Auto-refresh activation time / 自动刷新激活时间 ---
-        post["last_active"] = now_iso()
-
-        try:
-            with open(file_path, "w", encoding="utf-8") as f:
-                f.write(frontmatter.dumps(post))
-        except OSError as e:
-            logger.error(f"Failed to write bucket update / 写入桶更新失败: {file_path}: {e}")
-            return False
-
-        logger.info(f"Updated bucket / 更新记忆桶: {bucket_id}")
-        return True
-
-    # ---------------------------------------------------------
-    # Wikilink injection
-    # 自动添加 Obsidian 双链
-    # ---------------------------------------------------------
-    def _apply_wikilinks(
-        self,
-        content: str,
-        tags: list[str],
-        domain: list[str],
-        name: str,
-    ) -> str:
-        """
-        Auto-inject Obsidian wikilinks, avoiding double-wrapping existing [[...]].
-        自动添加 Obsidian 双链，避免重复包裹已有 [[...]]。
-        """
-        if not self.wikilink_enabled or not content:
-            return content
-
-        keywords = self._collect_wikilink_keywords(content, tags, domain, name)
-        if not keywords:
-            return content
-
-        # Split on existing wikilinks to avoid wrapping them again
-        # 按已有双链切分，避免重复包裹
-        segments = re.split(r"(\[\[[^\]]+\]\])", content)
-        pattern = re.compile("|".join(re.escape(kw) for kw in keywords))
-        for i, segment in enumerate(segments):
-            if segment.startswith("[[") and segment.endswith("]]"):
-                continue
-            updated = pattern.sub(lambda m: f"[[{m.group(0)}]]", segment)
-            segments[i] = updated
-        return "".join(segments)
-
-    def _collect_wikilink_keywords(
-        self,
-        content: str,
-        tags: list[str],
-        domain: list[str],
-        name: str,
-    ) -> list[str]:
-        """
-        Collect candidate keywords from tags/domain/auto-extraction.
-        汇总候选关键词：可选 tags/domain + 自动提词。
-        """
-        candidates = []
-
-        if self.wikilink_use_tags:
-            candidates.extend(tags or [])
-        if self.wikilink_use_domain:
-            candidates.extend(domain or [])
-        if name:
-            candidates.append(name)
-        if self.wikilink_use_auto_keywords:
-            candidates.extend(self._extract_auto_keywords(content))
-
-        return self._normalize_keywords(candidates)
-
-    def _normalize_keywords(self, keywords: list[str]) -> list[str]:
-        """
-        Deduplicate and sort by length (longer first to avoid short words
-        breaking long ones during replacement).
-        去重并按长度排序，优先替换长词。
-        """
-        if not keywords:
-            return []
-
-        seen = set()
-        cleaned = []
-        for keyword in keywords:
-            if not isinstance(keyword, str):
-                continue
-            kw = keyword.strip()
-            if len(kw) < self.wikilink_min_len:
-                continue
-            if kw in self.wikilink_exclude_keywords:
-                continue
-            if kw.lower() in self.wikilink_stopwords:
-                continue
-            if kw in seen:
-                continue
-            seen.add(kw)
-            cleaned.append(kw)
-
-        return sorted(cleaned, key=len, reverse=True)
-
-    def _extract_auto_keywords(self, content: str) -> list[str]:
-        """
-        Auto-extract keywords from body text, prioritizing high-frequency words.
-        从正文自动提词，优先高频词。
-        """
-        if not content:
-            return []
-
-        try:
-            zh_words = [w.strip() for w in jieba.lcut(content) if w.strip()]
-        except Exception:
-            zh_words = []
-        en_words = re.findall(r"[A-Za-z][A-Za-z0-9_-]{2,20}", content)
-
-        # Chinese bigrams / 中文双词组合
-        zh_bigrams = []
-        for i in range(len(zh_words) - 1):
-            left = zh_words[i]
-            right = zh_words[i + 1]
-            if len(left) < self.wikilink_min_len or len(right) < self.wikilink_min_len:
-                continue
-            if not re.fullmatch(r"[\u4e00-\u9fff]+", left + right):
-                continue
-            if len(left + right) > 8:
-                continue
-            zh_bigrams.append(left + right)
-
-        merged = []
-        for word in zh_words + zh_bigrams + en_words:
-            if len(word) < self.wikilink_min_len:
-                continue
-            if re.fullmatch(r"\d+", word):
-                continue
-            if word.lower() in self.wikilink_stopwords:
-                continue
-            merged.append(word)
-
-        if not merged:
-            return []
-
-        counter = Counter(merged)
-        return [w for w, _ in counter.most_common(self.wikilink_auto_top_k)]
-
-    # ---------------------------------------------------------
-    # Delete bucket
-    # 删除桶
-    # ---------------------------------------------------------
-    async def delete(self, bucket_id: str) -> bool:
-        """
-        Delete a memory bucket file.
-        删除指定的记忆桶文件。
-        """
-        file_path = self._find_bucket_file(bucket_id)
-        if not file_path:
-            return False
-
-        try:
-            os.remove(file_path)
-        except OSError as e:
-            logger.error(f"Failed to delete bucket file / 删除桶文件失败: {file_path}: {e}")
-            return False
-
-        logger.info(f"Deleted bucket / 删除记忆桶: {bucket_id}")
-        return True
-
-    # ---------------------------------------------------------
-    # Touch bucket (refresh activation time + increment count)
-    # 触碰桶（刷新激活时间 + 累加激活次数）
-    # Called on every recall hit; affects decay score.
-    # 每次检索命中时调用，影响衰减得分。
-    # ---------------------------------------------------------
-    async def touch(self, bucket_id: str) -> None:
-        """
-        Update a bucket's last activation time and count.
-        更新桶的最后激活时间和激活次数。
-        """
-        file_path = self._find_bucket_file(bucket_id)
-        if not file_path:
-            return
-
-        try:
-            post = frontmatter.load(file_path)
-            post["last_active"] = now_iso()
-            post["activation_count"] = post.get("activation_count", 0) + 1
-
-            with open(file_path, "w", encoding="utf-8") as f:
-                f.write(frontmatter.dumps(post))
-        except Exception as e:
-            logger.warning(f"Failed to touch bucket / 触碰桶失败: {bucket_id}: {e}")
-
-    # ---------------------------------------------------------
-    # Multi-dimensional search (core feature)
-    # 多维搜索（核心功能）
-    #
-    # Strategy: domain pre-filter → weighted multi-dim ranking
-    # 策略：主题域预筛 → 多维加权精排
-    #
-    # Ranking formula:
-    #   total = topic(×w_topic) + emotion(×w_emotion)
-    #           + time(×w_time) + importance(×w_importance)
-    #
-    # Per-dimension scores (normalized to 0~1):
-    #   topic     = rapidfuzz weighted match (name/tags/domain/body)
-    #   emotion   = 1 - Euclidean distance (query v/a vs bucket v/a)
-    #   time      = e^(-0.02 × days) (recent memories first)
-    #   importance = importance / 10
-    # ---------------------------------------------------------
-    async def search(
-        self,
-        query: str,
-        limit: int = None,
-        domain_filter: list[str] = None,
-        query_valence: float = None,
-        query_arousal: float = None,
-    ) -> list[dict]:
-        """
-        Multi-dimensional indexed search for memory buckets.
-        多维索引搜索记忆桶。
-
-        domain_filter: pre-filter by domain (None = search all)
-        query_valence/arousal: emotion coordinates for resonance scoring
-        """
-        if not query or not query.strip():
-            return []
-
-        limit = limit or self.max_results
-        all_buckets = await self.list_all(include_archive=False)
-
-        if not all_buckets:
-            return []
-
-        # --- Layer 1: domain pre-filter (fast scope reduction) ---
-        # --- 第一层：主题域预筛（快速缩小范围）---
-        if domain_filter:
-            filter_set = {d.lower() for d in domain_filter}
-            candidates = [
-                b for b in all_buckets
-                if {d.lower() for d in b["metadata"].get("domain", [])} & filter_set
-            ]
-            # Fall back to full search if pre-filter yields nothing
-            # 预筛为空则回退全量搜索
-            if not candidates:
-                candidates = all_buckets
-        else:
-            candidates = all_buckets
-
-        # --- Layer 2: weighted multi-dim ranking ---
-        # --- 第二层：多维加权精排 ---
-        scored = []
-        for bucket in candidates:
-            meta = bucket.get("metadata", {})
-
-            try:
-                # Dim 1: topic relevance (fuzzy text, 0~1)
-                topic_score = self._calc_topic_score(query, bucket)
-
-                # Dim 2: emotion resonance (coordinate distance, 0~1)
-                emotion_score = self._calc_emotion_score(
-                    query_valence, query_arousal, meta
-                )
-
-                # Dim 3: time proximity (exponential decay, 0~1)
-                time_score = self._calc_time_score(meta)
-
-                # Dim 4: importance (direct normalization)
-                importance_score = max(1, min(10, int(meta.get("importance", 5)))) / 10.0
-
-                # --- Weighted sum / 加权求和 ---
-                total = (
-                    topic_score * self.w_topic
-                    + emotion_score * self.w_emotion
-                    + time_score * self.w_time
-                    + importance_score * self.w_importance
-                )
-                # Normalize to 0~100 for readability
-                weight_sum = self.w_topic + self.w_emotion + self.w_time + self.w_importance
-                normalized = (total / weight_sum) * 100 if weight_sum > 0 else 0
-
-                # Resolved buckets get ranking penalty (but still reachable by keyword)
-                # 已解决的桶降权排序（但仍可被关键词激活）
-                if meta.get("resolved", False):
-                    normalized *= 0.3
-
-                if normalized >= self.fuzzy_threshold:
-                    bucket["score"] = round(normalized, 2)
-                    scored.append(bucket)
-            except Exception as e:
-                logger.warning(
-                    f"Scoring failed for bucket {bucket.get('id', '?')} / "
-                    f"桶评分失败: {e}"
-                )
-                continue
-
-        scored.sort(key=lambda x: x["score"], reverse=True)
-        return scored[:limit]
-
-    # ---------------------------------------------------------
-    # Topic relevance sub-score:
-    # name(×3) + domain(×2.5) + tags(×2) + body(×1)
-    # 文本相关性子分：桶名(×3) + 主题域(×2.5) + 标签(×2) + 正文(×1)
-    # ---------------------------------------------------------
-    def _calc_topic_score(self, query: str, bucket: dict) -> float:
-        """
-        Calculate text dimension relevance score (0~1).
-        计算文本维度的相关性得分。
-        """
-        meta = bucket.get("metadata", {})
-
-        name_score = fuzz.partial_ratio(query, meta.get("name", "")) * 3
-        domain_score = (
-            max(
-                (fuzz.partial_ratio(query, d) for d in meta.get("domain", [])),
-                default=0,
-            )
-            * 2.5
-        )
-        tag_score = (
-            max(
-                (fuzz.partial_ratio(query, tag) for tag in meta.get("tags", [])),
-                default=0,
-            )
-            * 2
-        )
-        content_score = fuzz.partial_ratio(query, bucket.get("content", "")[:500]) * 1
-
-        return (name_score + domain_score + tag_score + content_score) / (100 * 8.5)
-
-    # ---------------------------------------------------------
-    # Emotion resonance sub-score:
-    # Based on Russell circumplex Euclidean distance
-    # 情感共鸣子分：基于环形情感模型的欧氏距离
-    # No emotion in query → neutral 0.5 (doesn't affect ranking)
-    # ---------------------------------------------------------
-    def _calc_emotion_score(
-        self, q_valence: float, q_arousal: float, meta: dict
-    ) -> float:
-        """
-        Calculate emotion resonance score (0~1, closer = higher).
-        计算情感共鸣度（0~1，越近越高）。
-        """
-        if q_valence is None or q_arousal is None:
-            return 0.5  # No emotion coordinates → neutral / 无情感坐标时给中性分
-
-        try:
-            b_valence = float(meta.get("valence", 0.5))
-            b_arousal = float(meta.get("arousal", 0.3))
-        except (ValueError, TypeError):
-            return 0.5
-
-        # Euclidean distance, max sqrt(2) ≈ 1.414
-        dist = math.sqrt((q_valence - b_valence) ** 2 + (q_arousal - b_arousal) ** 2)
-        return max(0.0, 1.0 - dist / 1.414)
-
-    # ---------------------------------------------------------
-    # Time proximity sub-score:
-    # More recent activation → higher score
-    # 时间亲近子分：距上次激活越近分越高
-    # ---------------------------------------------------------
-    def _calc_time_score(self, meta: dict) -> float:
-        """
-        Calculate time proximity score (0~1, more recent = higher).
-        计算时间亲近度。
-        """
-        last_active_str = meta.get("last_active", meta.get("created", ""))
-        try:
-            last_active = datetime.fromisoformat(str(last_active_str))
-            days = max(0.0, (datetime.now() - last_active).total_seconds() / 86400)
-        except (ValueError, TypeError):
-            days = 30
-        return math.exp(-0.02 * days)
-
-    # ---------------------------------------------------------
-    # List all buckets
-    # 列出所有桶
-    # ---------------------------------------------------------
-    async def list_all(self, include_archive: bool = False) -> list[dict]:
-        """
-        Recursively walk directories (including domain subdirs), list all buckets.
-        递归遍历目录（含域子目录），列出所有记忆桶。
-        """
-        buckets = []
-
-        dirs = [self.permanent_dir, self.dynamic_dir]
-        if include_archive:
-            dirs.append(self.archive_dir)
-
-        for dir_path in dirs:
-            if not os.path.exists(dir_path):
-                continue
-            for root, _, files in os.walk(dir_path):
-                for filename in files:
-                    if not filename.endswith(".md"):
-                        continue
-                    file_path = os.path.join(root, filename)
-                    bucket = self._load_bucket(file_path)
-                    if bucket:
-                        buckets.append(bucket)
-
-        return buckets
-
-    # ---------------------------------------------------------
-    # Statistics (counts per category + total size)
-    # 统计信息（各分类桶数量 + 总体积）
-    # ---------------------------------------------------------
-    async def get_stats(self) -> dict:
-        """
-        Return memory bucket statistics (including domain subdirs).
-        返回记忆桶的统计数据。
-        """
-        stats = {
-            "permanent_count": 0,
-            "dynamic_count": 0,
-            "archive_count": 0,
-            "total_size_kb": 0.0,
-            "domains": {},
-        }
-
-        for subdir, key in [
-            (self.permanent_dir, "permanent_count"),
-            (self.dynamic_dir, "dynamic_count"),
-            (self.archive_dir, "archive_count"),
-        ]:
-            if not os.path.exists(subdir):
-                continue
-            for root, _, files in os.walk(subdir):
-                for f in files:
-                    if f.endswith(".md"):
-                        stats[key] += 1
-                        fpath = os.path.join(root, f)
-                        try:
-                            stats["total_size_kb"] += os.path.getsize(fpath) / 1024
-                        except OSError:
-                            pass
-                        # Per-domain counts / 每个域的桶数量
-                        domain_name = os.path.basename(root)
-                        if domain_name != os.path.basename(subdir):
-                            stats["domains"][domain_name] = stats["domains"].get(domain_name, 0) + 1
-
-        return stats
-
-    # ---------------------------------------------------------
-    # Archive bucket (move from permanent/dynamic into archive)
-    # 归档桶（从 permanent/dynamic 移入 archive）
-    # Called by decay engine to simulate "forgetting"
-    # 由衰减引擎调用，模拟"遗忘"
-    # ---------------------------------------------------------
-    async def archive(self, bucket_id: str) -> bool:
-        """
-        Move a bucket into the archive directory (preserving domain subdirs).
-        将指定桶移入归档目录（保留域子目录结构）。
-        """
-        file_path = self._find_bucket_file(bucket_id)
-        if not file_path:
-            return False
-
-        try:
-            # Read once, get domain info and update type / 一次性读取
-            post = frontmatter.load(file_path)
-            domain = post.get("domain", ["未分类"])
-            primary_domain = sanitize_name(domain[0]) if domain else "未分类"
-            archive_subdir = os.path.join(self.archive_dir, primary_domain)
-            os.makedirs(archive_subdir, exist_ok=True)
-
-            dest = safe_path(archive_subdir, os.path.basename(file_path))
-
-            # Update type marker then move file / 更新类型标记后移动文件
-            post["type"] = "archived"
-            with open(file_path, "w", encoding="utf-8") as f:
-                f.write(frontmatter.dumps(post))
-
-            # Use shutil.move for cross-filesystem safety
-            # 使用 shutil.move 保证跨文件系统安全
-            shutil.move(file_path, str(dest))
-        except Exception as e:
-            logger.error(
-                f"Failed to archive bucket / 归档桶失败: {bucket_id}: {e}"
-            )
-            return False
-
-        logger.info(f"Archived bucket / 归档记忆桶: {bucket_id} → archive/{primary_domain}/")
-        return True
-
-    # ---------------------------------------------------------
-    # Internal: find bucket file across all three directories
-    # 内部：在三个目录中查找桶文件
-    # ---------------------------------------------------------
-    def _find_bucket_file(self, bucket_id: str) -> Optional[str]:
-        """
-        Recursively search permanent/dynamic/archive for a bucket file
-        matching the given ID.
-        在 permanent/dynamic/archive 中递归查找指定 ID 的桶文件。
-        """
-        if not bucket_id:
-            return None
-        for dir_path in [self.permanent_dir, self.dynamic_dir, self.archive_dir]:
-            if not os.path.exists(dir_path):
-                continue
-            for root, _, files in os.walk(dir_path):
-                for fname in files:
-                    if not fname.endswith(".md"):
-                        continue
-                    # Match by exact ID segment in filename
-                    # 通过文件名中的 ID 片段精确匹配
-                    if bucket_id in fname:
-                        return os.path.join(root, fname)
-        return None
-
-    # ---------------------------------------------------------
-    # Internal: load bucket data from .md file
-    # 内部：从 .md 文件加载桶数据
-    # ---------------------------------------------------------
-    def _load_bucket(self, file_path: str) -> Optional[dict]:
-        """
-        Parse a Markdown file and return structured bucket data.
-        解析 Markdown 文件，返回桶的结构化数据。
-        """
-        try:
-            post = frontmatter.load(file_path)
-            return {
-                "id": post.get("id", Path(file_path).stem),
-                "metadata": dict(post.metadata),
-                "content": post.content,
-                "path": file_path,
-            }
-        except Exception as e:
-            logger.warning(
-                f"Failed to load bucket file / 加载桶文件失败: {file_path}: {e}"
-            )
-            return None

backup_20260405_2124/decay_engine.py

@@ -1,242 +0,0 @@
-# ============================================================
-# Module: Memory Decay Engine (decay_engine.py)
-# 模块：记忆衰减引擎
-#
-# Simulates human forgetting curve; auto-decays inactive memories and archives them.
-# 模拟人类遗忘曲线，自动衰减不活跃记忆并归档。
-#
-# Core formula (improved Ebbinghaus + emotion coordinates):
-# 核心公式（改进版艾宾浩斯遗忘曲线 + 情感坐标）：
-#   Score = Importance × (activation_count^0.3) × e^(-λ×days) × emotion_weight
-#
-# Emotion weight (continuous coordinate, not discrete labels):
-# 情感权重（基于连续坐标而非离散列举）：
-#   emotion_weight = base + (arousal × arousal_boost)
-#   Higher arousal → higher emotion weight → slower decay
-#   唤醒度越高 → 情感权重越大 → 记忆衰减越慢
-#
-# Depended on by: server.py
-# 被谁依赖：server.py
-# ============================================================
-
-import math
-import asyncio
-import logging
-from datetime import datetime
-
-logger = logging.getLogger("ombre_brain.decay")
-
-
-class DecayEngine:
-    """
-    Memory decay engine — periodically scans all dynamic buckets,
-    calculates decay scores, auto-archives low-activity buckets
-    to simulate natural forgetting.
-    记忆衰减引擎 —— 定期扫描所有动态桶，
-    计算衰减得分，将低活跃桶自动归档，模拟自然遗忘。
-    """
-
-    def __init__(self, config: dict, bucket_mgr):
-        # --- Load decay parameters / 加载衰减参数 ---
-        decay_cfg = config.get("decay", {})
-        self.decay_lambda = decay_cfg.get("lambda", 0.05)
-        self.threshold = decay_cfg.get("threshold", 0.3)
-        self.check_interval = decay_cfg.get("check_interval_hours", 24)
-
-        # --- Emotion weight params (continuous arousal coordinate) ---
-        # --- 情感权重参数（基于连续 arousal 坐标）---
-        emotion_cfg = decay_cfg.get("emotion_weights", {})
-        self.emotion_base = emotion_cfg.get("base", 1.0)
-        self.arousal_boost = emotion_cfg.get("arousal_boost", 0.8)
-
-        self.bucket_mgr = bucket_mgr
-
-        # --- Background task control / 后台任务控制 ---
-        self._task: asyncio.Task | None = None
-        self._running = False
-
-    @property
-    def is_running(self) -> bool:
-        """Whether the decay engine is running in the background.
-        衰减引擎是否正在后台运行。"""
-        return self._running
-
-    # ---------------------------------------------------------
-    # Core: calculate decay score for a single bucket
-    # 核心：计算单个桶的衰减得分
-    #
-    # Higher score = more vivid memory; below threshold → archive
-    # 得分越高 = 记忆越鲜活，低于阈值则归档
-    # Permanent buckets never decay / 固化桶永远不衰减
-    # ---------------------------------------------------------
-    def calculate_score(self, metadata: dict) -> float:
-        """
-        Calculate current activity score for a memory bucket.
-        计算一个记忆桶的当前活跃度得分。
-
-        Formula: Score = Importance × (act_count^0.3) × e^(-λ×days) × (base + arousal×boost)
-        """
-        if not isinstance(metadata, dict):
-            return 0.0
-
-        # --- Permanent buckets never decay / 固化桶永不衰减 ---
-        if metadata.get("type") == "permanent":
-            return 999.0
-
-        importance = max(1, min(10, int(metadata.get("importance", 5))))
-        activation_count = max(1, int(metadata.get("activation_count", 1)))
-
-        # --- Days since last activation / 距离上次激活过了多少天 ---
-        last_active_str = metadata.get("last_active", metadata.get("created", ""))
-        try:
-            last_active = datetime.fromisoformat(str(last_active_str))
-            days_since = max(0.0, (datetime.now() - last_active).total_seconds() / 86400)
-        except (ValueError, TypeError):
-            days_since = 30  # Parse failure → assume 30 days / 解析失败假设已过 30 天
-
-        # --- Emotion weight: continuous arousal coordinate ---
-        # --- 情感权重：基于连续 arousal 坐标计算 ---
-        # Higher arousal → stronger emotion → higher weight → slower decay
-        # arousal 越高 → 情感越强烈 → 权重越大 → 衰减越慢
-        try:
-            arousal = max(0.0, min(1.0, float(metadata.get("arousal", 0.3))))
-        except (ValueError, TypeError):
-            arousal = 0.3
-        emotion_weight = self.emotion_base + arousal * self.arousal_boost
-
-        # --- Apply decay formula / 套入衰减公式 ---
-        score = (
-            importance
-            * (activation_count ** 0.3)
-            * math.exp(-self.decay_lambda * days_since)
-            * emotion_weight
-        )
-
-        # --- Weight pool modifiers / 权重池修正因子 ---
-        # Resolved events drop to 5%, sink to bottom awaiting keyword reactivation
-        # 已解决的事件权重骤降到 5%，沉底等待关键词激活
-        resolved_factor = 0.05 if metadata.get("resolved", False) else 1.0
-        # High-arousal unresolved buckets get urgency boost for priority surfacing
-        # 高唤醒未解决桶额外加成，优先浮现
-        urgency_boost = 1.5 if (arousal > 0.7 and not metadata.get("resolved", False)) else 1.0
-
-        return round(score * resolved_factor * urgency_boost, 4)
-
-    # ---------------------------------------------------------
-    # Execute one decay cycle
-    # 执行一轮衰减周期
-    # Scan all dynamic buckets → score → archive those below threshold
-    # 扫描所有动态桶 → 算分 → 低于阈值的归档
-    # ---------------------------------------------------------
-    async def run_decay_cycle(self) -> dict:
-        """
-        Execute one decay cycle: iterate dynamic buckets, archive those
-        scoring below threshold.
-        执行一轮衰减：遍历动态桶，归档得分低于阈值的桶。
-
-        Returns stats: {"checked": N, "archived": N, "lowest_score": X}
-        """
-        try:
-            buckets = await self.bucket_mgr.list_all(include_archive=False)
-        except Exception as e:
-            logger.error(f"Failed to list buckets for decay / 衰减周期列桶失败: {e}")
-            return {"checked": 0, "archived": 0, "lowest_score": 0, "error": str(e)}
-
-        checked = 0
-        archived = 0
-        lowest_score = float("inf")
-
-        for bucket in buckets:
-            meta = bucket.get("metadata", {})
-
-            # Skip permanent buckets / 跳过固化桶
-            if meta.get("type") == "permanent":
-                continue
-
-            checked += 1
-            try:
-                score = self.calculate_score(meta)
-            except Exception as e:
-                logger.warning(
-                    f"Score calculation failed for {bucket.get('id', '?')} / "
-                    f"计算得分失败: {e}"
-                )
-                continue
-
-            lowest_score = min(lowest_score, score)
-
-            # --- Below threshold → archive (simulate forgetting) ---
-            # --- 低于阈值 → 归档（模拟遗忘）---
-            if score < self.threshold:
-                try:
-                    success = await self.bucket_mgr.archive(bucket["id"])
-                    if success:
-                        archived += 1
-                        logger.info(
-                            f"Decay archived / 衰减归档: "
-                            f"{meta.get('name', bucket['id'])} "
-                            f"(score={score:.4f}, threshold={self.threshold})"
-                        )
-                except Exception as e:
-                    logger.warning(
-                        f"Archive failed for {bucket.get('id', '?')} / "
-                        f"归档失败: {e}"
-                    )
-
-        result = {
-            "checked": checked,
-            "archived": archived,
-            "lowest_score": lowest_score if checked > 0 else 0,
-        }
-        logger.info(f"Decay cycle complete / 衰减周期完成: {result}")
-        return result
-
-    # ---------------------------------------------------------
-    # Background decay task management
-    # 后台衰减任务管理
-    # ---------------------------------------------------------
-    async def ensure_started(self) -> None:
-        """
-        Ensure the decay engine is started (lazy init on first call).
-        确保衰减引擎已启动（懒加载，首次调用时启动）。
-        """
-        if not self._running:
-            await self.start()
-
-    async def start(self) -> None:
-        """Start the background decay loop.
-        启动后台衰减循环。"""
-        if self._running:
-            return
-        self._running = True
-        self._task = asyncio.create_task(self._background_loop())
-        logger.info(
-            f"Decay engine started, interval: {self.check_interval}h / "
-            f"衰减引擎已启动，检查间隔: {self.check_interval} 小时"
-        )
-
-    async def stop(self) -> None:
-        """Stop the background decay loop.
-        停止后台衰减循环。"""
-        self._running = False
-        if self._task:
-            self._task.cancel()
-            try:
-                await self._task
-            except asyncio.CancelledError:
-                pass
-        logger.info("Decay engine stopped / 衰减引擎已停止")
-
-    async def _background_loop(self) -> None:
-        """Background loop: run decay → sleep → repeat.
-        后台循环体：执行衰减 → 睡眠 → 重复。"""
-        while self._running:
-            try:
-                await self.run_decay_cycle()
-            except Exception as e:
-                logger.error(f"Decay cycle error / 衰减周期出错: {e}")
-            # --- Wait for next cycle / 等待下一个周期 ---
-            try:
-                await asyncio.sleep(self.check_interval * 3600)
-            except asyncio.CancelledError:
-                break

backup_20260405_2124/server.py

@@ -1,536 +0,0 @@
-# ============================================================
-# Module: MCP Server Entry Point (server.py)
-# 模块：MCP 服务器主入口
-#
-# Starts the Ombre Brain MCP service and registers memory
-# operation tools for Claude to call.
-# 启动 Ombre Brain MCP 服务，注册记忆操作工具供 Claude 调用。
-#
-# Core responsibilities:
-# 核心职责：
-#   - Initialize config, bucket manager, dehydrator, decay engine
-#     初始化配置、记忆桶管理器、脱水器、衰减引擎
-#   - Expose 5 MCP tools:
-#     暴露 5 个 MCP 工具：
-#       breath — Surface unresolved memories or search by keyword
-#                浮现未解决记忆 或 按关键词检索
-#       hold   — Store a single memory
-#                存储单条记忆
-#       grow   — Diary digest, auto-split into multiple buckets
-#                日记归档，自动拆分多桶
-#       trace  — Modify metadata / resolved / delete
-#                修改元数据 / resolved 标记 / 删除
-#       pulse  — System status + bucket listing
-#                系统状态 + 所有桶列表
-#
-# Startup:
-# 启动方式：
-#   Local:  python server.py
-#   Remote: OMBRE_TRANSPORT=streamable-http python server.py
-#   Docker: docker-compose up
-# ============================================================
-
-import os
-import sys
-import random
-import logging
-import asyncio
-import httpx
-
-# --- Ensure same-directory modules can be imported ---
-# --- 确保同目录下的模块能被正确导入 ---
-sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
-
-from mcp.server.fastmcp import FastMCP
-
-from bucket_manager import BucketManager
-from dehydrator import Dehydrator
-from decay_engine import DecayEngine
-from utils import load_config, setup_logging
-
-# --- Load config & init logging / 加载配置 & 初始化日志 ---
-config = load_config()
-setup_logging(config.get("log_level", "INFO"))
-logger = logging.getLogger("ombre_brain")
-
-# --- Initialize three core components / 初始化三大核心组件 ---
-bucket_mgr = BucketManager(config)                  # Bucket manager / 记忆桶管理器
-dehydrator = Dehydrator(config)                      # Dehydrator / 脱水器
-decay_engine = DecayEngine(config, bucket_mgr)       # Decay engine / 衰减引擎
-
-# --- Create MCP server instance / 创建 MCP 服务器实例 ---
-# host="0.0.0.0" so Docker container's SSE is externally reachable
-# stdio mode ignores host (no network)
-mcp = FastMCP(
-    "Ombre Brain",
-    host="0.0.0.0",
-    port=8000,
-)
-
-
-# =============================================================
-# /health endpoint: lightweight keepalive
-# 轻量保活接口
-# For Cloudflare Tunnel or reverse proxy to ping, preventing idle timeout
-# 供 Cloudflare Tunnel 或反代定期 ping，防止空闲超时断连
-# =============================================================
-@mcp.custom_route("/health", methods=["GET"])
-async def health_check(request):
-    from starlette.responses import JSONResponse
-    try:
-        stats = await bucket_mgr.get_stats()
-        return JSONResponse({
-            "status": "ok",
-            "buckets": stats["permanent_count"] + stats["dynamic_count"],
-            "decay_engine": "running" if decay_engine.is_running else "stopped",
-        })
-    except Exception as e:
-        return JSONResponse({"status": "error", "detail": str(e)}, status_code=500)
-
-
-# =============================================================
-# Internal helper: merge-or-create
-# 内部辅助：检查是否可合并，可以则合并，否则新建
-# Shared by hold and grow to avoid duplicate logic
-# hold 和 grow 共用，避免重复逻辑
-# =============================================================
-async def _merge_or_create(
-    content: str,
-    tags: list,
-    importance: int,
-    domain: list,
-    valence: float,
-    arousal: float,
-    name: str = "",
-) -> tuple[str, bool]:
-    """
-    Check if a similar bucket exists for merging; merge if so, create if not.
-    Returns (bucket_id_or_name, is_merged).
-    检查是否有相似桶可合并，有则合并，无则新建。
-    返回 (桶ID或名称, 是否合并)。
-    """
-    try:
-        existing = await bucket_mgr.search(content, limit=1)
-    except Exception as e:
-        logger.warning(f"Search for merge failed, creating new / 合并搜索失败，新建: {e}")
-        existing = []
-
-    if existing and existing[0].get("score", 0) > config.get("merge_threshold", 75):
-        bucket = existing[0]
-        try:
-            merged = await dehydrator.merge(bucket["content"], content)
-            await bucket_mgr.update(
-                bucket["id"],
-                content=merged,
-                tags=list(set(bucket["metadata"].get("tags", []) + tags)),
-                importance=max(bucket["metadata"].get("importance", 5), importance),
-                domain=list(set(bucket["metadata"].get("domain", []) + domain)),
-                valence=valence,
-                arousal=arousal,
-            )
-            return bucket["metadata"].get("name", bucket["id"]), True
-        except Exception as e:
-            logger.warning(f"Merge failed, creating new / 合并失败，新建: {e}")
-
-    bucket_id = await bucket_mgr.create(
-        content=content,
-        tags=tags,
-        importance=importance,
-        domain=domain,
-        valence=valence,
-        arousal=arousal,
-        name=name or None,
-    )
-    return bucket_id, False
-
-
-# =============================================================
-# Tool 1: breath — Breathe
-# 工具 1：breath — 呼吸
-#
-# No args: surface highest-weight unresolved memories (active push)
-# 无参数：浮现权重最高的未解决记忆
-# With args: search by keyword + emotion coordinates
-# 有参数：按关键词+情感坐标检索记忆
-# =============================================================
-@mcp.tool()
-async def breath(
-    query: str = "",
-    max_results: int = 3,
-    domain: str = "",
-    valence: float = -1,
-    arousal: float = -1,
-) -> str:
-    """检索记忆或浮现未解决记忆。query 为空时自动推送权重最高的未解决桶；有 query 时按关键词+情感检索。domain 逗号分隔，valence/arousal 传 0~1 启用情感共鸣，-1 忽略。"""
-    await decay_engine.ensure_started()
-
-    # --- No args: surfacing mode (weight pool active push) ---
-    # --- 无参数：浮现模式（权重池主动推送）---
-    if not query.strip():
-        try:
-            all_buckets = await bucket_mgr.list_all(include_archive=False)
-        except Exception as e:
-            logger.error(f"Failed to list buckets for surfacing / 浮现列桶失败: {e}")
-            return "记忆系统暂时无法访问。"
-
-        unresolved = [
-            b for b in all_buckets
-            if not b["metadata"].get("resolved", False)
-            and b["metadata"].get("type") != "permanent"
-        ]
-        if not unresolved:
-            return "权重池平静，没有需要处理的记忆。"
-
-        scored = sorted(
-            unresolved,
-            key=lambda b: decay_engine.calculate_score(b["metadata"]),
-            reverse=True,
-        )
-        top = scored[:2]
-        results = []
-        for b in top:
-            try:
-                summary = await dehydrator.dehydrate(b["content"], b["metadata"])
-                await bucket_mgr.touch(b["id"])
-                score = decay_engine.calculate_score(b["metadata"])
-                results.append(f"[权重:{score:.2f}] {summary}")
-            except Exception as e:
-                logger.warning(f"Failed to dehydrate surfaced bucket / 浮现脱水失败: {e}")
-                continue
-        if not results:
-            return "权重池平静，没有需要处理的记忆。"
-        return "=== 浮现记忆 ===\n" + "\n---\n".join(results)
-
-    # --- With args: search mode / 有参数：检索模式 ---
-    domain_filter = [d.strip() for d in domain.split(",") if d.strip()] or None
-    q_valence = valence if 0 <= valence <= 1 else None
-    q_arousal = arousal if 0 <= arousal <= 1 else None
-
-    try:
-        matches = await bucket_mgr.search(
-            query,
-            limit=max_results,
-            domain_filter=domain_filter,
-            query_valence=q_valence,
-            query_arousal=q_arousal,
-        )
-    except Exception as e:
-        logger.error(f"Search failed / 检索失败: {e}")
-        return "检索过程出错，请稍后重试。"
-
-    results = []
-    for bucket in matches:
-        try:
-            summary = await dehydrator.dehydrate(bucket["content"], bucket["metadata"])
-            await bucket_mgr.touch(bucket["id"])
-            results.append(summary)
-        except Exception as e:
-            logger.warning(f"Failed to dehydrate search result / 检索结果脱水失败: {e}")
-            continue
-
-    # --- Random surfacing: when search returns < 3, 40% chance to float old memories ---
-    # --- 随机浮现：检索结果不足 3 条时，40% 概率从低权重旧桶里漂上来 ---
-    if len(matches) < 3 and random.random() < 0.4:
-        try:
-            all_buckets = await bucket_mgr.list_all(include_archive=False)
-            matched_ids = {b["id"] for b in matches}
-            low_weight = [
-                b for b in all_buckets
-                if b["id"] not in matched_ids
-                and decay_engine.calculate_score(b["metadata"]) < 2.0
-            ]
-            if low_weight:
-                drifted = random.sample(low_weight, min(random.randint(1, 3), len(low_weight)))
-                drift_results = []
-                for b in drifted:
-                    summary = await dehydrator.dehydrate(b["content"], b["metadata"])
-                    drift_results.append(f"[surface_type: random]\n{summary}")
-                results.append("--- 忽然想起来 ---\n" + "\n---\n".join(drift_results))
-        except Exception as e:
-            logger.warning(f"Random surfacing failed / 随机浮现失败: {e}")
-
-    if not results:
-        return "未找到相关记忆。"
-
-    return "\n---\n".join(results)
-
-
-# =============================================================
-# Tool 2: hold — Hold on to this
-# 工具 2：hold — 握住，留下来
-# =============================================================
-@mcp.tool()
-async def hold(
-    content: str,
-    tags: str = "",
-    importance: int = 5,
-) -> str:
-    """存储单条记忆。自动打标+合并相似桶。tags 逗号分隔，importance 1-10。"""
-    await decay_engine.ensure_started()
-
-    # --- Input validation / 输入校验 ---
-    if not content or not content.strip():
-        return "内容为空，无法存储。"
-
-    importance = max(1, min(10, importance))
-    extra_tags = [t.strip() for t in tags.split(",") if t.strip()]
-
-    # --- Step 1: auto-tagging / 自动打标 ---
-    try:
-        analysis = await dehydrator.analyze(content)
-    except Exception as e:
-        logger.warning(f"Auto-tagging failed, using defaults / 自动打标失败: {e}")
-        analysis = {
-            "domain": ["未分类"], "valence": 0.5, "arousal": 0.3,
-            "tags": [], "suggested_name": "",
-        }
-
-    domain = analysis["domain"]
-    valence = analysis["valence"]
-    arousal = analysis["arousal"]
-    auto_tags = analysis["tags"]
-    suggested_name = analysis.get("suggested_name", "")
-
-    all_tags = list(dict.fromkeys(auto_tags + extra_tags))
-
-    # --- Step 2: merge or create / 合并或新建 ---
-    result_name, is_merged = await _merge_or_create(
-        content=content,
-        tags=all_tags,
-        importance=importance,
-        domain=domain,
-        valence=valence,
-        arousal=arousal,
-        name=suggested_name,
-    )
-
-    if is_merged:
-        return (
-            f"已合并到现有记忆桶: {result_name}\n"
-            f"主题域: {', '.join(domain)} | 情感: V{valence:.1f}/A{arousal:.1f}"
-        )
-    return (
-        f"已创建新记忆桶: {result_name}\n"
-        f"主题域: {', '.join(domain)} | 情感: V{valence:.1f}/A{arousal:.1f} | 标签: {', '.join(all_tags)}"
-    )
-
-
-# =============================================================
-# Tool 3: grow — Grow, fragments become memories
-# 工具 3：grow — 生长，一天的碎片长成记忆
-# =============================================================
-@mcp.tool()
-async def grow(content: str) -> str:
-    """日记归档。自动拆分长内容为多个记忆桶。"""
-    await decay_engine.ensure_started()
-
-    if not content or not content.strip():
-        return "内容为空，无法整理。"
-
-    # --- Step 1: let API split and organize / 让 API 拆分整理 ---
-    try:
-        items = await dehydrator.digest(content)
-    except Exception as e:
-        logger.error(f"Diary digest failed / 日记整理失败: {e}")
-        return f"日记整理失败: {e}"
-
-    if not items:
-        return "内容为空或整理失败。"
-
-    results = []
-    created = 0
-    merged = 0
-
-    # --- Step 2: merge or create each item (with per-item error handling) ---
-    # --- 逐条合并或新建（单条失败不影响其他）---
-    for item in items:
-        try:
-            result_name, is_merged = await _merge_or_create(
-                content=item["content"],
-                tags=item.get("tags", []),
-                importance=item.get("importance", 5),
-                domain=item.get("domain", ["未分类"]),
-                valence=item.get("valence", 0.5),
-                arousal=item.get("arousal", 0.3),
-                name=item.get("name", ""),
-            )
-
-            if is_merged:
-                results.append(f"  📎 合并 → {result_name}")
-                merged += 1
-            else:
-                domains_str = ",".join(item.get("domain", []))
-                results.append(
-                    f"  📝 新建 [{item.get('name', result_name)}] "
-                    f"主题:{domains_str} V{item.get('valence', 0.5):.1f}/A{item.get('arousal', 0.3):.1f}"
-                )
-                created += 1
-        except Exception as e:
-            logger.warning(
-                f"Failed to process diary item / 日记条目处理失败: "
-                f"{item.get('name', '?')}: {e}"
-            )
-            results.append(f"  ⚠️ 失败: {item.get('name', '未知条目')}")
-
-    summary = f"=== 日记整理完成 ===\n拆分为 {len(items)} 条 | 新建 {created} 桶 | 合并 {merged} 桶\n"
-    return summary + "\n".join(results)
-
-
-# =============================================================
-# Tool 4: trace — Trace, redraw the outline of a memory
-# 工具 4：trace — 描摹，重新勾勒记忆的轮廓
-# Also handles deletion (delete=True)
-# 同时承接删除功能
-# =============================================================
-@mcp.tool()
-async def trace(
-    bucket_id: str,
-    name: str = "",
-    domain: str = "",
-    valence: float = -1,
-    arousal: float = -1,
-    importance: int = -1,
-    tags: str = "",
-    resolved: int = -1,
-    delete: bool = False,
-) -> str:
-    """修改记忆元数据。resolved=1 标记已解决（桶权重骤降沉底），resolved=0 重新激活，delete=True 删除桶。其余字段只传需改的，-1 或空串表示不改。"""
-
-    if not bucket_id or not bucket_id.strip():
-        return "请提供有效的 bucket_id。"
-
-    # --- Delete mode / 删除模式 ---
-    if delete:
-        success = await bucket_mgr.delete(bucket_id)
-        return f"已遗忘记忆桶: {bucket_id}" if success else f"未找到记忆桶: {bucket_id}"
-
-    bucket = await bucket_mgr.get(bucket_id)
-    if not bucket:
-        return f"未找到记忆桶: {bucket_id}"
-
-    # --- Collect only fields actually passed / 只收集用户实际传入的字段 ---
-    updates = {}
-    if name:
-        updates["name"] = name
-    if domain:
-        updates["domain"] = [d.strip() for d in domain.split(",") if d.strip()]
-    if 0 <= valence <= 1:
-        updates["valence"] = valence
-    if 0 <= arousal <= 1:
-        updates["arousal"] = arousal
-    if 1 <= importance <= 10:
-        updates["importance"] = importance
-    if tags:
-        updates["tags"] = [t.strip() for t in tags.split(",") if t.strip()]
-    if resolved in (0, 1):
-        updates["resolved"] = bool(resolved)
-
-    if not updates:
-        return "没有任何字段需要修改。"
-
-    success = await bucket_mgr.update(bucket_id, **updates)
-    if not success:
-        return f"修改失败: {bucket_id}"
-
-    changed = ", ".join(f"{k}={v}" for k, v in updates.items())
-    # Explicit hint about resolved state change semantics
-    # 特别提示 resolved 状态变化的语义
-    if "resolved" in updates:
-        if updates["resolved"]:
-            changed += " → 已沉底，只在关键词触发时重新浮现"
-        else:
-            changed += " → 已重新激活，将参与浮现排序"
-    return f"已修改记忆桶 {bucket_id}: {changed}"
-
-
-# =============================================================
-# Tool 5: pulse — Heartbeat, system status + memory listing
-# 工具 5：pulse — 脉搏，系统状态 + 记忆列表
-# =============================================================
-@mcp.tool()
-async def pulse(include_archive: bool = False) -> str:
-    """系统状态和所有记忆桶摘要。include_archive=True 时包含归档桶。"""
-    try:
-        stats = await bucket_mgr.get_stats()
-    except Exception as e:
-        return f"获取系统状态失败: {e}"
-
-    status = (
-        f"=== Ombre Brain 记忆系统 ===\n"
-        f"固化记忆桶: {stats['permanent_count']} 个\n"
-        f"动态记忆桶: {stats['dynamic_count']} 个\n"
-        f"归档记忆桶: {stats['archive_count']} 个\n"
-        f"总存储大小: {stats['total_size_kb']:.1f} KB\n"
-        f"衰减引擎: {'运行中' if decay_engine.is_running else '已停止'}\n"
-    )
-
-    # --- List all bucket summaries / 列出所有桶摘要 ---
-    try:
-        buckets = await bucket_mgr.list_all(include_archive=include_archive)
-    except Exception as e:
-        return status + f"\n列出记忆桶失败: {e}"
-
-    if not buckets:
-        return status + "\n记忆库为空。"
-
-    lines = []
-    for b in buckets:
-        meta = b.get("metadata", {})
-        if meta.get("type") == "permanent":
-            icon = "📦"
-        elif meta.get("type") == "archived":
-            icon = "🗄️"
-        elif meta.get("resolved", False):
-            icon = "✅"
-        else:
-            icon = "💭"
-        try:
-            score = decay_engine.calculate_score(meta)
-        except Exception:
-            score = 0.0
-        domains = ",".join(meta.get("domain", []))
-        val = meta.get("valence", 0.5)
-        aro = meta.get("arousal", 0.3)
-        resolved_tag = " [已解决]" if meta.get("resolved", False) else ""
-        lines.append(
-            f"{icon} [{meta.get('name', b['id'])}]{resolved_tag} "
-            f"主题:{domains} "
-            f"情感:V{val:.1f}/A{aro:.1f} "
-            f"重要:{meta.get('importance', '?')} "
-            f"权重:{score:.2f} "
-            f"标签:{','.join(meta.get('tags', []))}"
-        )
-
-    return status + "\n=== 记忆列表 ===\n" + "\n".join(lines)
-
-
-# --- Entry point / 启动入口 ---
-if __name__ == "__main__":
-    transport = config.get("transport", "stdio")
-    logger.info(f"Ombre Brain starting | transport: {transport}")
-
-    # --- Application-level keepalive: remote mode only, ping /health every 60s ---
-    # --- 应用层保活：仅远程模式下启动，每 60 秒 ping 一次 /health ---
-    # Prevents Cloudflare Tunnel from dropping idle connections
-    if transport in ("sse", "streamable-http"):
-        async def _keepalive_loop():
-            await asyncio.sleep(10)  # Wait for server to fully start
-            async with httpx.AsyncClient() as client:
-                while True:
-                    try:
-                        await client.get("http://localhost:8000/health", timeout=5)
-                        logger.debug("Keepalive ping OK / 保活 ping 成功")
-                    except Exception as e:
-                        logger.warning(f"Keepalive ping failed / 保活 ping 失败: {e}")
-                    await asyncio.sleep(60)
-
-        import threading
-
-        def _start_keepalive():
-            loop = asyncio.new_event_loop()
-            loop.run_until_complete(_keepalive_loop())
-
-        t = threading.Thread(target=_start_keepalive, daemon=True)
-        t.start()
-
-    mcp.run(transport=transport)

bucket_manager.py

@@ -28,15 +28,12 @@
 import os
 import math
 import logging
-import re
 import shutil
-from collections import Counter
 from datetime import datetime
 from pathlib import Path
 from typing import Optional
 
 import frontmatter
-import jieba
 from rapidfuzz import fuzz
 
 from utils import generate_bucket_id, sanitize_name, safe_path, now_iso
@@ -54,7 +51,7 @@ class BucketManager:
     天然兼容 Obsidian 直接浏览和编辑。
     """
 
-    def __init__(self, config: dict):
+    def __init__(self, config: dict, embedding_engine=None):
         # --- Read storage paths from config / 从配置中读取存储路径 ---
         self.base_dir = config["buckets_dir"]
         self.permanent_dir = os.path.join(self.base_dir, "permanent")
@@ -88,9 +85,12 @@ class BucketManager:
         scoring = config.get("scoring_weights", {})
         self.w_topic = scoring.get("topic_relevance", 4.0)
         self.w_emotion = scoring.get("emotion_resonance", 2.0)
-        self.w_time = scoring.get("time_proximity", 2.5)
+        self.w_time = scoring.get("time_proximity", 1.5)
         self.w_importance = scoring.get("importance", 1.0)
-        self.content_weight = scoring.get("content_weight", 3.0)  # Added to allow better content-based matching during merge
+        self.content_weight = scoring.get("content_weight", 1.0)  # body×1, per spec
+
+        # --- Optional embedding engine for pre-filtering / 可选 embedding 引擎，用于预筛候选集 ---
+        self.embedding_engine = embedding_engine
 
     # ---------------------------------------------------------
     # Create a new bucket
@@ -121,6 +121,10 @@ class BucketManager:
         """
         bucket_id = generate_bucket_id()
         bucket_name = sanitize_name(name) if name else bucket_id
+        # feel buckets are allowed to have empty domain; others default to ["未分类"]
+        if bucket_type == "feel":
+            domain = domain if domain is not None else []
+        else:
             domain = domain or ["未分类"]
         tags = tags or []
         linked_content = content  # wikilink injection disabled; LLM adds [[]] via prompt
@@ -142,7 +146,7 @@ class BucketManager:
             "type": bucket_type,
             "created": now_iso(),
             "last_active": now_iso(),
-            "activation_count": 1,
+            "activation_count": 0,
         }
         if pinned:
             metadata["pinned"] = True
@@ -289,19 +293,17 @@ class BucketManager:
             logger.error(f"Failed to write bucket update / 写入桶更新失败: {file_path}: {e}")
             return False
 
-        # --- Auto-move: pinned → permanent/, resolved → archive/ ---
+        # --- Auto-move: pinned → permanent/ ---
-        # --- 自动移动：钉选 → permanent/，已解决 → archive/ ---
+        # --- 自动移动：钉选 → permanent/ ---
+        # NOTE: resolved buckets are NOT auto-archived here.
+        # They stay in dynamic/ and decay naturally until score < threshold.
+        # 注意：resolved 桶不在此自动归档，留在 dynamic/ 随衰减引擎自然归档。
         domain = post.get("domain", ["未分类"])
         if kwargs.get("pinned") and post.get("type") != "permanent":
             post["type"] = "permanent"
             with open(file_path, "w", encoding="utf-8") as f:
                 f.write(frontmatter.dumps(post))
             self._move_bucket(file_path, self.permanent_dir, domain)
-        elif kwargs.get("resolved") and post.get("type") not in ("permanent", "feel"):
-            post["type"] = "archived"
-            with open(file_path, "w", encoding="utf-8") as f:
-                f.write(frontmatter.dumps(post))
-            self._move_bucket(file_path, self.archive_dir, domain)
 
         logger.info(f"Updated bucket / 更新记忆桶: {bucket_id}")
         return True
@@ -473,6 +475,20 @@ class BucketManager:
         else:
             candidates = all_buckets
 
+        # --- Layer 1.5: embedding pre-filter (optional, reduces multi-dim ranking set) ---
+        # --- 第1.5层：embedding 预筛（可选，缩小精排候选集）---
+        if self.embedding_engine and self.embedding_engine.enabled:
+            try:
+                vector_results = await self.embedding_engine.search_similar(query, top_k=50)
+                if vector_results:
+                    vector_ids = {bid for bid, _ in vector_results}
+                    emb_candidates = [b for b in candidates if b["id"] in vector_ids]
+                    if emb_candidates:  # only replace if there's non-empty overlap
+                        candidates = emb_candidates
+                    # else: keep original candidates as fallback
+            except Exception as e:
+                logger.warning(f"Embedding pre-filter failed, using fuzzy only / embedding 预筛失败: {e}")
+
         # --- Layer 2: weighted multi-dim ranking ---
         # --- 第二层：多维加权精排 ---
         scored = []
@@ -505,12 +521,14 @@ class BucketManager:
                 weight_sum = self.w_topic + self.w_emotion + self.w_time + self.w_importance
                 normalized = (total / weight_sum) * 100 if weight_sum > 0 else 0
 
+                # Threshold check uses raw (pre-penalty) score so resolved buckets
+                # 阈值用原始分数判定，确保 resolved 桶在关键词命中时仍可被搜出
+                # remain reachable by keyword (penalty applied only to ranking).
+                if normalized >= self.fuzzy_threshold:
                     # Resolved buckets get ranking penalty (but still reachable by keyword)
-                # 已解决的桶降权排序（但仍可被关键词激活）
+                    # 已解决的桶仅在排序时降权
                     if meta.get("resolved", False):
                         normalized *= 0.3
-
-                if normalized >= self.fuzzy_threshold:
                     bucket["score"] = round(normalized, 2)
                     scored.append(bucket)
             except Exception as e:
@@ -596,7 +614,7 @@ class BucketManager:
             days = max(0.0, (datetime.now() - last_active).total_seconds() / 86400)
         except (ValueError, TypeError):
             days = 30
-        return math.exp(-0.1 * days)
+        return math.exp(-0.02 * days)
 
     # ---------------------------------------------------------
     # List all buckets

check_icloud_conflicts.py

@@ -0,0 +1,147 @@
+#!/usr/bin/env python3
+# ============================================================
+# check_icloud_conflicts.py  —  Ombre Brain iCloud Conflict Detector
+# iCloud 冲突文件检测器
+#
+# Scans the configured bucket directory for iCloud sync conflict
+# artefacts and duplicate bucket IDs, then prints a report.
+# 扫描配置的桶目录，发现 iCloud 同步冲突文件及重复桶 ID，输出报告。
+#
+# Usage:
+#   python check_icloud_conflicts.py
+#   python check_icloud_conflicts.py --buckets-dir /path/to/dir
+#   python check_icloud_conflicts.py --quiet      # exit-code only (0=clean)
+# ============================================================
+
+from __future__ import annotations
+
+import argparse
+import os
+import re
+import sys
+from collections import defaultdict
+from pathlib import Path
+
+# ──────────────────────────────────────────────────────────────
+# iCloud conflict file patterns
+# Pattern 1 (macOS classic): "filename 2.md", "filename 3.md"
+# Pattern 2 (iCloud Drive):  "filename (Device's conflicted copy YYYY-MM-DD).md"
+# ──────────────────────────────────────────────────────────────
+_CONFLICT_SUFFIX   = re.compile(r"^(.+?)\s+\d+\.md$")
+_CONFLICT_ICLOUD   = re.compile(r"^(.+?)\s+\(.+conflicted copy .+\)\.md$", re.IGNORECASE)
+# Bucket ID pattern: 12 hex chars at end of stem before extension
+_BUCKET_ID_PATTERN = re.compile(r"_([0-9a-f]{12})$")
+
+
+def resolve_buckets_dir() -> Path:
+    """Resolve bucket directory: env var → config.yaml → ./buckets fallback."""
+    env_dir = os.environ.get("OMBRE_BUCKETS_DIR", "").strip()
+    if env_dir:
+        return Path(env_dir)
+
+    config_path = Path(__file__).parent / "config.yaml"
+    if config_path.exists():
+        try:
+            import yaml  # type: ignore
+            with open(config_path, encoding="utf-8") as f:
+                cfg = yaml.safe_load(f) or {}
+            if cfg.get("buckets_dir"):
+                return Path(cfg["buckets_dir"])
+        except Exception:
+            pass
+
+    return Path(__file__).parent / "buckets"
+
+
+def scan(buckets_dir: Path) -> tuple[list[Path], dict[str, list[Path]]]:
+    """
+    Returns:
+        conflict_files  — list of files that look like iCloud conflict artefacts
+        dup_ids         — dict of bucket_id -> [list of files sharing that id]
+                          (only entries with 2+ files)
+    """
+    if not buckets_dir.exists():
+        return [], {}
+
+    conflict_files: list[Path] = []
+    id_to_files: dict[str, list[Path]] = defaultdict(list)
+
+    for md_file in buckets_dir.rglob("*.md"):
+        name = md_file.name
+
+        # --- Conflict file detection ---
+        if _CONFLICT_SUFFIX.match(name) or _CONFLICT_ICLOUD.match(name):
+            conflict_files.append(md_file)
+            continue  # don't register conflicts in the ID map
+
+        # --- Duplicate ID detection ---
+        stem = md_file.stem
+        m = _BUCKET_ID_PATTERN.search(stem)
+        if m:
+            id_to_files[m.group(1)].append(md_file)
+
+    dup_ids = {bid: paths for bid, paths in id_to_files.items() if len(paths) > 1}
+    return conflict_files, dup_ids
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser(
+        description="Detect iCloud conflict files and duplicate bucket IDs."
+    )
+    parser.add_argument(
+        "--buckets-dir",
+        metavar="PATH",
+        help="Override bucket directory (default: from config.yaml / OMBRE_BUCKETS_DIR)",
+    )
+    parser.add_argument(
+        "--quiet",
+        action="store_true",
+        help="Suppress output; exit 0 = clean, 1 = problems found",
+    )
+    args = parser.parse_args()
+
+    buckets_dir = Path(args.buckets_dir) if args.buckets_dir else resolve_buckets_dir()
+
+    if not args.quiet:
+        print(f"Scanning: {buckets_dir}")
+        if not buckets_dir.exists():
+            print("  ✗ Directory does not exist.")
+            return 1
+        print()
+
+    conflict_files, dup_ids = scan(buckets_dir)
+    problems = bool(conflict_files or dup_ids)
+
+    if args.quiet:
+        return 1 if problems else 0
+
+    # ── Report ─────────────────────────────────────────────────
+    if not problems:
+        print("✓ No iCloud conflicts or duplicate IDs found.")
+        return 0
+
+    if conflict_files:
+        print(f"⚠ iCloud conflict files ({len(conflict_files)} found):")
+        for f in sorted(conflict_files):
+            rel = f.relative_to(buckets_dir) if f.is_relative_to(buckets_dir) else f
+            print(f"  {rel}")
+        print()
+
+    if dup_ids:
+        print(f"⚠ Duplicate bucket IDs ({len(dup_ids)} ID(s) shared by multiple files):")
+        for bid, paths in sorted(dup_ids.items()):
+            print(f"  ID: {bid}")
+            for p in sorted(paths):
+                rel = p.relative_to(buckets_dir) if p.is_relative_to(buckets_dir) else p
+                print(f"    {rel}")
+        print()
+
+    print(
+        "NOTE: This script is report-only. No files are modified or deleted.\n"
+        "注意：本脚本仅报告，不删除或修改任何文件。"
+    )
+    return 1
+
+
+if __name__ == "__main__":
+    sys.exit(main())

config.example.yaml

@@ -28,9 +28,11 @@ log_level: "INFO"
 merge_threshold: 75
 
 # --- Dehydration API / 脱水压缩 API 配置 ---
-# Uses a cheap LLM for intelligent compression; auto-degrades to local
+# Uses a cheap LLM for intelligent compression. API is required; if the
-# keyword extraction if API is unavailable
+# configured key/endpoint is unavailable, hold/grow will raise an explicit
-# 用廉价 LLM 做智能压缩，API 不可用时自动降级到本地关键词提取
+# error instead of silently degrading (see BEHAVIOR_SPEC.md 三、降级行为表).
+# 用廉价 LLM 做智能压缩。API 为必需；如 key/endpoint 不可用，
+# hold/grow 会直接报错而非静默降级（详见 BEHAVIOR_SPEC.md 三、降级行为表）。
 dehydration:
   # Supports any OpenAI-compatible API: DeepSeek / Ollama / LM Studio / vLLM / Gemini etc.
   # 支持所有 OpenAI 兼容 API：DeepSeek / Ollama / LM Studio / vLLM / Gemini 等
@@ -61,11 +63,14 @@ decay:
 # --- Embedding / 向量化配置 ---
 # Uses embedding API for semantic similarity search
 # 通过 embedding API 实现语义相似度搜索
-# Reuses the same API key (OMBRE_API_KEY) and base_url from dehydration config
+# You can configure embedding independently from dehydration.
-# 复用脱水配置的 API key 和 base_url
+# If api_key is omitted, reuses the same API key (OMBRE_API_KEY) and base_url from dehydration config
+# 你可以把 embedding 独立配置；若 api_key 留空，复用脱水配置的 API key 和 base_url
 embedding:
   enabled: true                   # Enable embedding / 启用向量化
   model: "gemini-embedding-001"   # Embedding model / 向量化模型
+  # base_url: "https://generativelanguage.googleapis.com/v1beta/openai"
+  # api_key: ""                  
 
 # --- Scoring weights / 检索权重参数 ---
 # total = topic(×4) + emotion(×2) + time(×1.5) + importance(×1)

dashboard.html

@@ -607,6 +607,7 @@
   <div class="search-bar">
     <input type="text" id="search-input" placeholder="搜索记忆…" />
   </div>
+  <button onclick="doLogout()" title="退出登录" style="margin-left:12px;background:none;border:1px solid var(--border);color:var(--text-dim);border-radius:20px;padding:6px 14px;font-size:12px;cursor:pointer;">退出</button>
 </div>
 
 <div class="tabs">
@@ -615,6 +616,7 @@
   <div class="tab" data-tab="network">记忆网络</div>
   <div class="tab" data-tab="config">配置</div>
   <div class="tab" data-tab="import">导入</div>
+  <div class="tab" data-tab="settings">设置</div>
 </div>
 
 <div class="content" id="list-view">
@@ -778,7 +780,259 @@
   <div id="detail-content"></div>
 </div>
 
+<!-- Settings Tab View -->
+<div class="content" id="settings-view" style="display:none">
+  <div style="max-width:580px;margin:0 auto;">
+
+    <div class="config-section">
+      <h3>服务状态</h3>
+      <div id="settings-status" style="font-size:13px;color:var(--text-dim);line-height:2;">加载中…</div>
+      <button onclick="loadSettingsStatus()" style="margin-top:8px;font-size:12px;padding:4px 12px;">刷新状态</button>
+    </div>
+
+    <div class="config-section">
+      <h3>修改密码</h3>
+      <div id="settings-env-notice" style="display:none;font-size:12px;color:var(--warning);margin-bottom:10px;">
+        ⚠ 当前使用环境变量 OMBRE_DASHBOARD_PASSWORD，请直接修改环境变量。
+      </div>
+      <div id="settings-pwd-form">
+        <div class="config-row">
+          <label>当前密码</label>
+          <input type="password" id="settings-current-pwd" placeholder="当前密码" />
+        </div>
+        <div class="config-row">
+          <label>新密码</label>
+          <input type="password" id="settings-new-pwd" placeholder="新密码（至少6位）" />
+        </div>
+        <div class="config-row">
+          <label>确认新密码</label>
+          <input type="password" id="settings-new-pwd2" placeholder="再次输入新密码" />
+        </div>
+        <button class="btn-primary" onclick="changePassword()" style="margin-top:4px;">修改密码</button>
+        <div id="settings-pwd-msg" style="margin-top:10px;font-size:13px;"></div>
+      </div>
+    </div>
+
+    <div class="config-section">
+      <h3>宿主机记忆桶目录 (Docker)</h3>
+      <div style="font-size:12px;color:var(--text-dim);margin-bottom:10px;line-height:1.6;">
+        设置 docker-compose 中 <code>${OMBRE_HOST_VAULT_DIR:-./buckets}:/data</code> 的宿主机路径。
+        留空则使用项目内 <code>./buckets</code>。
+        <span style="color:var(--warning);">⚠ 修改后需在宿主机执行 <code>docker compose down && docker compose up -d</code> 才会生效。</span>
+      </div>
+      <div class="config-row">
+        <label>路径</label>
+        <input type="text" id="settings-host-vault" placeholder="例如 /Users/you/Obsidian/Ombre Brain" style="flex:1;" />
+      </div>
+      <div style="display:flex;gap:8px;align-items:center;margin-top:6px;">
+        <button class="btn-primary" onclick="saveHostVault()">保存到 .env</button>
+        <button onclick="loadHostVault()" style="font-size:12px;padding:4px 12px;">重新加载</button>
+        <span id="settings-host-vault-msg" style="font-size:12px;"></span>
+      </div>
+    </div>
+
+    <div class="config-section">
+      <h3>账号操作</h3>
+      <button onclick="doLogout()" style="color:var(--negative);border-color:var(--negative);">退出登录</button>
+    </div>
+
+  </div>
+</div>
+
+<!-- Auth Overlay -->
+<div id="auth-overlay" style="position:fixed;inset:0;z-index:9999;background:var(--bg-gradient);background-attachment:fixed;display:flex;align-items:center;justify-content:center;">
+  <div style="background:var(--surface);backdrop-filter:blur(16px);border:1px solid var(--border);border-radius:24px;padding:48px 40px;max-width:400px;width:90%;text-align:center;box-shadow:0 8px 40px rgba(0,0,0,0.12);">
+    <h2 style="font-family:'Cormorant Garamond',serif;font-size:28px;color:var(--accent);margin-bottom:8px;">◐ Ombre Brain</h2>
+    <p style="color:var(--text-dim);font-size:13px;margin-bottom:28px;" id="auth-subtitle">验证身份</p>
+
+    <!-- Setup form -->
+    <div id="auth-setup-form" style="display:none;">
+      <p style="font-size:13px;color:var(--text-dim);margin-bottom:16px;">首次使用，请设置访问密码</p>
+      <input type="password" id="auth-setup-pwd" placeholder="设置密码（至少6位）" style="display:block;width:100%;padding:10px 16px;border:1px solid var(--border);border-radius:10px;background:var(--surface-solid);color:var(--text);font-size:14px;margin-bottom:10px;" />
+      <input type="password" id="auth-setup-pwd2" placeholder="确认密码" style="display:block;width:100%;padding:10px 16px;border:1px solid var(--border);border-radius:10px;background:var(--surface-solid);color:var(--text);font-size:14px;margin-bottom:16px;" onkeydown="if(event.key==='Enter')doSetup()" />
+      <button onclick="doSetup()" style="width:100%;padding:12px;background:var(--accent);color:#fff;border:none;border-radius:10px;font-size:14px;cursor:pointer;">设置密码并进入</button>
+    </div>
+
+    <!-- Login form -->
+    <div id="auth-login-form" style="display:none;">
+      <input type="password" id="auth-login-pwd" placeholder="输入访问密码" style="display:block;width:100%;padding:10px 16px;border:1px solid var(--border);border-radius:10px;background:var(--surface-solid);color:var(--text);font-size:14px;margin-bottom:16px;" onkeydown="if(event.key==='Enter')doLogin()" />
+      <button onclick="doLogin()" style="width:100%;padding:12px;background:var(--accent);color:#fff;border:none;border-radius:10px;font-size:14px;cursor:pointer;">登录</button>
+    </div>
+
+    <div id="auth-error" style="color:var(--negative);font-size:13px;margin-top:12px;display:none;"></div>
+  </div>
+</div>
+
 <script>
+// ========================================
+// Auth system / 认证系统
+// ========================================
+async function checkAuth() {
+  try {
+    const resp = await fetch('/auth/status');
+    const data = await resp.json();
+    if (data.setup_needed) {
+      document.getElementById('auth-subtitle').textContent = '首次设置';
+      document.getElementById('auth-setup-form').style.display = 'block';
+    } else if (data.authenticated) {
+      document.getElementById('auth-overlay').style.display = 'none';
+    } else {
+      document.getElementById('auth-subtitle').textContent = '请输入访问密码';
+      document.getElementById('auth-login-form').style.display = 'block';
+    }
+  } catch {
+    document.getElementById('auth-overlay').style.display = 'none';
+  }
+}
+
+function showAuthError(msg) {
+  const el = document.getElementById('auth-error');
+  el.textContent = msg;
+  el.style.display = 'block';
+}
+
+async function doSetup() {
+  const p1 = document.getElementById('auth-setup-pwd').value;
+  const p2 = document.getElementById('auth-setup-pwd2').value;
+  if (p1.length < 6) return showAuthError('密码至少6位');
+  if (p1 !== p2) return showAuthError('两次密码不一致');
+  const resp = await fetch('/auth/setup', { method: 'POST', headers: {'Content-Type':'application/json'}, body: JSON.stringify({password: p1}) });
+  if (resp.ok) {
+    document.getElementById('auth-overlay').style.display = 'none';
+  } else {
+    const d = await resp.json();
+    showAuthError(d.detail || '设置失败');
+  }
+}
+
+async function doLogin() {
+  const pwd = document.getElementById('auth-login-pwd').value;
+  const resp = await fetch('/auth/login', { method: 'POST', headers: {'Content-Type':'application/json'}, body: JSON.stringify({password: pwd}) });
+  if (resp.ok) {
+    document.getElementById('auth-overlay').style.display = 'none';
+  } else {
+    const d = await resp.json();
+    showAuthError(d.detail || '密码错误');
+  }
+}
+
+async function doLogout() {
+  await fetch('/auth/logout', { method: 'POST' });
+  document.getElementById('auth-setup-form').style.display = 'none';
+  document.getElementById('auth-login-form').style.display = 'none';
+  document.getElementById('auth-login-form').style.display = 'block';
+  document.getElementById('auth-subtitle').textContent = '请输入访问密码';
+  document.getElementById('auth-error').style.display = 'none';
+  document.getElementById('auth-overlay').style.display = 'flex';
+}
+
+async function changePassword() {
+  const currentPwd = document.getElementById('settings-current-pwd').value;
+  const newPwd = document.getElementById('settings-new-pwd').value;
+  const newPwd2 = document.getElementById('settings-new-pwd2').value;
+  const msgEl = document.getElementById('settings-pwd-msg');
+  if (newPwd.length < 6) { msgEl.style.color = 'var(--negative)'; msgEl.textContent = '新密码至少6位'; return; }
+  if (newPwd !== newPwd2) { msgEl.style.color = 'var(--negative)'; msgEl.textContent = '两次密码不一致'; return; }
+  const resp = await authFetch('/auth/change-password', { method: 'POST', headers: {'Content-Type':'application/json'}, body: JSON.stringify({current: currentPwd, new: newPwd}) });
+  if (!resp) return;
+  if (resp.ok) {
+    msgEl.style.color = 'var(--accent)'; msgEl.textContent = '密码修改成功';
+    document.getElementById('settings-current-pwd').value = '';
+    document.getElementById('settings-new-pwd').value = '';
+    document.getElementById('settings-new-pwd2').value = '';
+  } else {
+    const d = await resp.json();
+    msgEl.style.color = 'var(--negative)'; msgEl.textContent = d.detail || '修改失败';
+  }
+}
+
+async function loadSettingsStatus() {
+  const el = document.getElementById('settings-status');
+  try {
+    const resp = await authFetch('/api/status');
+    if (!resp) return;
+    const d = await resp.json();
+    const noticeEl = document.getElementById('settings-env-notice');
+    if (d.using_env_password) noticeEl.style.display = 'block';
+    else noticeEl.style.display = 'none';
+    el.innerHTML = `
+      <b>版本</b>：${d.version}<br>
+      <b>Bucket 总数</b>：${(d.buckets?.total ?? 0)} （永久:${d.buckets?.permanent ?? 0} / 动态:${d.buckets?.dynamic ?? 0} / 归档:${d.buckets?.archive ?? 0}）<br>
+      <b>衰减引擎</b>：${d.decay_engine}<br>
+      <b>向量搜索</b>：${d.embedding_enabled ? '已启用' : '未启用'}<br>
+    `;
+  } catch(e) {
+    el.textContent = '加载失败: ' + e;
+  }
+  // Also refresh the host-vault input whenever the settings tab is loaded.
+  loadHostVault();
+}
+
+async function loadHostVault() {
+  const input = document.getElementById('settings-host-vault');
+  const msg = document.getElementById('settings-host-vault-msg');
+  if (!input) return;
+  msg.textContent = '';
+  msg.style.color = 'var(--text-dim)';
+  try {
+    const resp = await authFetch('/api/host-vault');
+    if (!resp) return;
+    const d = await resp.json();
+    input.value = d.value || '';
+    if (d.source === 'env') {
+      msg.textContent = '当前由进程环境变量提供（修改 .env 不会立即覆盖）';
+      msg.style.color = 'var(--warning)';
+    } else if (d.source === 'file') {
+      msg.textContent = '当前来自 ' + (d.env_file || '.env');
+    } else {
+      msg.textContent = '尚未设置（默认使用 ./buckets）';
+    }
+  } catch(e) {
+    msg.style.color = 'var(--negative)';
+    msg.textContent = '加载失败: ' + e;
+  }
+}
+
+async function saveHostVault() {
+  const input = document.getElementById('settings-host-vault');
+  const msg = document.getElementById('settings-host-vault-msg');
+  if (!input) return;
+  const value = input.value.trim();
+  msg.textContent = '保存中…';
+  msg.style.color = 'var(--text-dim)';
+  try {
+    const resp = await authFetch('/api/host-vault', {
+      method: 'POST',
+      headers: {'Content-Type': 'application/json'},
+      body: JSON.stringify({value})
+    });
+    if (!resp) return;
+    const d = await resp.json();
+    if (resp.ok) {
+      msg.style.color = 'var(--accent)';
+      msg.textContent = '已保存 → ' + (d.env_file || '.env') + '（需重启容器生效）';
+    } else {
+      msg.style.color = 'var(--negative)';
+      msg.textContent = d.error || '保存失败';
+    }
+  } catch(e) {
+    msg.style.color = 'var(--negative)';
+    msg.textContent = '保存失败: ' + e;
+  }
+}
+
+// authFetch: wraps fetch, shows auth overlay on 401
+async function authFetch(url, options) {
+  const resp = await fetch(url, options);
+  if (resp.status === 401) {
+    doLogout();
+    return null;
+  }
+  return resp;
+}
+
+// ========================================
+
 const BASE = location.origin;
 let allBuckets = [];
 let currentFilter = 'all';
@@ -793,9 +1047,11 @@ document.querySelectorAll('.tab').forEach(tab => {
     document.getElementById('network-view').style.display = target === 'network' ? '' : 'none';
     document.getElementById('config-view').style.display = target === 'config' ? '' : 'none';
     document.getElementById('import-view').style.display = target === 'import' ? '' : 'none';
+    document.getElementById('settings-view').style.display = target === 'settings' ? '' : 'none';
     if (target === 'network') loadNetwork();
     if (target === 'config') loadConfig();
     if (target === 'import') { pollImportStatus(); loadImportResults(); }
+    if (target === 'settings') loadSettingsStatus();
   });
 });
 
@@ -812,7 +1068,11 @@ document.getElementById('search-input').addEventListener('input', (e) => {
 async function loadBuckets() {
   try {
     const res = await fetch(BASE + '/api/buckets');
-    allBuckets = await res.json();
+    const data = await res.json();
+    if (!res.ok || !Array.isArray(data)) {
+      throw new Error((data && data.error) ? data.error : `HTTP ${res.status}`);
+    }
+    allBuckets = data;
     updateStats();
     buildFilters();
     renderBuckets(allBuckets);
@@ -1237,7 +1497,7 @@ async function saveConfig(persist) {
   }
 }
 
-loadBuckets();
+checkAuth().then(() => loadBuckets());
 
 // --- Import functions ---
 const uploadZone = document.getElementById('import-upload-zone');
@@ -1300,6 +1560,7 @@ function updateImportUI(s) {
   document.getElementById('import-status-text').textContent = statusMap[s.status] || s.status;
   document.getElementById('import-pause-btn').style.display = s.status === 'running' ? '' : 'none';
   if (s.status !== 'running') clearInterval(importPollTimer);
+  if (s.status === 'completed') loadImportResults();
   const errDiv = document.getElementById('import-errors');
   if (s.errors && s.errors.length) {
     errDiv.style.display = '';

decay_engine.py

@@ -112,7 +112,7 @@ class DecayEngine:
             return 50.0
 
         importance = max(1, min(10, int(metadata.get("importance", 5))))
-        activation_count = max(1, int(metadata.get("activation_count", 1)))
+        activation_count = max(1.0, float(metadata.get("activation_count", 1)))
 
         # --- Days since last activation ---
         last_active_str = metadata.get("last_active", metadata.get("created", ""))
@@ -215,6 +215,7 @@ class DecayEngine:
                 if imp <= 4 and days_since > 30:
                     try:
                         await self.bucket_mgr.update(bucket["id"], resolved=True)
+                        meta["resolved"] = True  # refresh local meta so resolved_factor applies this cycle
                         auto_resolved += 1
                         logger.info(
                             f"Auto-resolved / 自动结案: "

dehydrator.py

@@ -152,10 +152,13 @@ class Dehydrator:
     """
     Data dehydrator + content analyzer.
     Three capabilities: dehydration / merge / auto-tagging (domain + emotion).
-    Prefers API (better quality); auto-degrades to local (guaranteed availability).
+    API-only: every public method requires a working LLM API.
+    If the API is unavailable, methods raise RuntimeError so callers can
+    surface the failure to the user instead of silently producing low-quality results.
     数据脱水器 + 内容分析器。
     三大能力：脱水压缩 / 新旧合并 / 自动打标。
-    优先走 API，API 挂了自动降级到本地。
+    仅走 API：API 不可用时直接抛出 RuntimeError，调用方明确感知。
+    （根据 BEHAVIOR_SPEC.md 三、降级行为表决策：无本地降级）
     """
 
     def __init__(self, config: dict):
@@ -235,8 +238,8 @@ class Dehydrator:
     # ---------------------------------------------------------
     # Dehydrate: compress raw content into concise summary
     # 脱水：将原始内容压缩为精简摘要
-    # Try API first, fallback to local
+    # API only (no local fallback)
-    # 先尝试 API，失败则回退本地
+    # 仅通过 API 脱水（无本地回退）
     # ---------------------------------------------------------
     async def dehydrate(self, content: str, metadata: dict = None) -> str:
         """

docker-compose.user.yml

@@ -19,7 +19,20 @@ services:
       - OMBRE_API_KEY=${OMBRE_API_KEY}
       - OMBRE_TRANSPORT=streamable-http
       - OMBRE_BUCKETS_DIR=/data
+      # --- Model override (optional) ---
+      # If you use Gemini instead of DeepSeek, set these in your .env:
+      # 如使用 Gemini 而非 DeepSeek，在 .env 里加：
+      #   OMBRE_DEHYDRATION_BASE_URL=https://generativelanguage.googleapis.com/v1beta/openai/
+      #   OMBRE_DEHYDRATION_MODEL=gemini-2.5-flash-lite
+      #   OMBRE_EMBEDDING_BASE_URL=https://generativelanguage.googleapis.com/v1beta/openai/
+      - OMBRE_DEHYDRATION_BASE_URL=${OMBRE_DEHYDRATION_BASE_URL:-}
+      - OMBRE_DEHYDRATION_MODEL=${OMBRE_DEHYDRATION_MODEL:-}
+      - OMBRE_EMBEDDING_BASE_URL=${OMBRE_EMBEDDING_BASE_URL:-}
+      - OMBRE_EMBEDDING_MODEL=${OMBRE_EMBEDDING_MODEL:-}
     volumes:
       # 改成你的 Obsidian Vault 路径，或保持 ./buckets 用本地目录
       # Change to your Obsidian Vault path, or keep ./buckets for local storage
       - ./buckets:/data
+      # (Optional) Mount custom config to override model / API settings:
+      # （可选）挂载自定义配置，覆盖模型和 API 设置：
+      # - ./config.yaml:/app/config.yaml

docker-compose.yml

@@ -21,11 +21,17 @@ services:
       - OMBRE_TRANSPORT=streamable-http     # Claude.ai requires streamable-http
       - OMBRE_BUCKETS_DIR=/data             # Container-internal bucket path / 容器内路径
     volumes:
-      # Mount your Obsidian vault (or any host directory) for persistent storage
+      # Mount your Obsidian vault (or any host directory) for persistent storage.
-      # 挂载你的 Obsidian 仓库（或任意宿主机目录）做持久化存储
+      # Set OMBRE_HOST_VAULT_DIR in your .env (or in the Dashboard "Storage" panel)
-      # Example / 示例:
+      # to point at the host folder you want mounted into the container at /data.
-      #   - /path/to/your/Obsidian Vault/Ombre Brain:/data
+      # 挂载你的 Obsidian 仓库（或任意宿主机目录）做持久化存储。
-      - /Users/p0lar1s/Library/Mobile Documents/iCloud~md~obsidian/Documents/Obsidian Vault/Ombre Brain:/data
+      # 在 .env（或 Dashboard 的「存储」面板）中设置 OMBRE_HOST_VAULT_DIR
+      # 指向你希望挂载到容器 /data 的宿主机目录。
+      #
+      # Examples / 示例:
+      #   OMBRE_HOST_VAULT_DIR=/path/to/your/Obsidian Vault/Ombre Brain
+      #   OMBRE_HOST_VAULT_DIR=~/Library/Mobile Documents/iCloud~md~obsidian/Documents/Obsidian Vault/Ombre Brain
+      - ${OMBRE_HOST_VAULT_DIR:-./buckets}:/data
       - ./config.yaml:/app/config.yaml
 
   # Cloudflare Tunnel (optional) — expose to public internet

embedding_engine.py

@@ -16,8 +16,6 @@ import json
 import math
 import sqlite3
 import logging
-import asyncio
-from pathlib import Path
 
 from openai import AsyncOpenAI
 
@@ -34,8 +32,12 @@ class EmbeddingEngine:
         dehy_cfg = config.get("dehydration", {})
         embed_cfg = config.get("embedding", {})
 
-        self.api_key = dehy_cfg.get("api_key", "")
+        self.api_key = (embed_cfg.get("api_key") or dehy_cfg.get("api_key") or "").strip()
-        self.base_url = dehy_cfg.get("base_url", "https://generativelanguage.googleapis.com/v1beta/openai/")
+        self.base_url = (
+            (embed_cfg.get("base_url") or "").strip()
+            or (dehy_cfg.get("base_url") or "").strip()
+            or "https://generativelanguage.googleapis.com/v1beta/openai/"
+        )
         self.model = embed_cfg.get("model", "gemini-embedding-001")
         self.enabled = bool(self.api_key) and embed_cfg.get("enabled", True)
 

import_memory.py

@@ -19,10 +19,8 @@ import os
 import json
 import hashlib
 import logging
-import asyncio
 from datetime import datetime
 from pathlib import Path
-from typing import Optional
 
 from utils import count_tokens_approx, now_iso
 
@@ -39,6 +37,8 @@ def _parse_claude_json(data: dict | list) -> list[dict]:
     turns = []
     conversations = data if isinstance(data, list) else [data]
     for conv in conversations:
+        if not isinstance(conv, dict):
+            continue
         messages = conv.get("chat_messages", conv.get("messages", []))
         for msg in messages:
             if not isinstance(msg, dict):
@@ -61,18 +61,27 @@ def _parse_chatgpt_json(data: list | dict) -> list[dict]:
     turns = []
     conversations = data if isinstance(data, list) else [data]
     for conv in conversations:
+        if not isinstance(conv, dict):
+            continue
         mapping = conv.get("mapping", {})
         if mapping:
             # ChatGPT uses a tree structure with mapping
-            sorted_nodes = sorted(
+            # Filter out None nodes before sorting
-                mapping.values(),
+            valid_nodes = [n for n in mapping.values() if isinstance(n, dict)]
-                key=lambda n: n.get("message", {}).get("create_time", 0) or 0,
+
-            )
+            def _node_ts(n):
+                msg = n.get("message")
+                if not isinstance(msg, dict):
+                    return 0
+                return msg.get("create_time") or 0
+
+            sorted_nodes = sorted(valid_nodes, key=_node_ts)
             for node in sorted_nodes:
                 msg = node.get("message")
                 if not msg or not isinstance(msg, dict):
                     continue
-                content_parts = msg.get("content", {}).get("parts", [])
+                content_obj = msg.get("content", {})
+                content_parts = content_obj.get("parts", []) if isinstance(content_obj, dict) else []
                 content = " ".join(str(p) for p in content_parts if p)
                 if not content.strip():
                     continue
@@ -168,7 +177,7 @@ def detect_and_parse(raw_content: str, filename: str = "") -> list[dict]:
                 # Single conversation object with role/content messages
                 if "role" in sample and "content" in sample:
                     return _parse_claude_json(data)
-        except (json.JSONDecodeError, KeyError, IndexError):
+        except (json.JSONDecodeError, KeyError, IndexError, AttributeError, TypeError):
             pass
 
     # Fall back to markdown/text

migrate_to_domains.py

@@ -12,7 +12,25 @@ import os
 import re
 import shutil
 
-VAULT_DIR = os.path.expanduser("~/Documents/Obsidian Vault/Ombre Brain")
+
+def _resolve_vault_dir() -> str:
+    """
+    Resolve the bucket vault root.
+    Priority: $OMBRE_BUCKETS_DIR > config.yaml > built-in ./buckets.
+    """
+    env_dir = os.environ.get("OMBRE_BUCKETS_DIR", "").strip()
+    if env_dir:
+        return os.path.expanduser(env_dir)
+    try:
+        from utils import load_config
+        return load_config()["buckets_dir"]
+    except Exception:
+        return os.path.join(
+            os.path.dirname(os.path.abspath(__file__)), "buckets"
+        )
+
+
+VAULT_DIR = _resolve_vault_dir()
 DYNAMIC_DIR = os.path.join(VAULT_DIR, "dynamic")
 
 
@@ -99,7 +117,7 @@ def migrate():
         print(f"  ✓ {filename}")
         print(f"    → {primary_domain}/{new_filename}")
 
-    print(f"\n迁移完成。")
+    print("\n迁移完成。")
 
     # 展示新结构
     print("\n=== 新目录结构 ===")

reclassify_api.py

@@ -38,7 +38,11 @@ ANALYZE_PROMPT = (
     '}'
 )
 
-DATA_DIR = "/data/dynamic"
+DATA_DIR = os.path.join(
+    os.environ.get("OMBRE_BUCKETS_DIR", "").strip()
+    or (lambda: __import__("utils").load_config()["buckets_dir"])(),
+    "dynamic",
+)
 UNCLASS_DIR = os.path.join(DATA_DIR, "未分类")
 
 
@@ -48,11 +52,15 @@ def sanitize(name):
 
 
 async def reclassify():
+    from utils import load_config
+    cfg = load_config()
+    dehy = cfg.get("dehydration", {})
     client = AsyncOpenAI(
-        api_key=os.environ.get("OMBRE_API_KEY", ""),
+        api_key=os.environ.get("OMBRE_API_KEY", "") or dehy.get("api_key", ""),
-        base_url="https://api.siliconflow.cn/v1",
+        base_url=dehy.get("base_url", "https://api.deepseek.com/v1"),
         timeout=60.0,
     )
+    model_name = dehy.get("model", "deepseek-chat")
 
     files = sorted(glob.glob(os.path.join(UNCLASS_DIR, "*.md")))
     print(f"找到 {len(files)} 个未分类文件\n")
@@ -66,7 +74,7 @@ async def reclassify():
 
         try:
             resp = await client.chat.completions.create(
-                model="deepseek-ai/DeepSeek-V3",
+                model=model_name,
                 messages=[
                     {"role": "system", "content": ANALYZE_PROMPT},
                     {"role": "user", "content": full_text[:2000]},

reclassify_domains.py

@@ -8,7 +8,25 @@ import os
 import re
 import shutil
 
-VAULT_DIR = os.path.expanduser("~/Documents/Obsidian Vault/Ombre Brain")
+
+def _resolve_vault_dir() -> str:
+    """
+    Resolve the bucket vault root.
+    Priority: $OMBRE_BUCKETS_DIR > config.yaml > built-in ./buckets.
+    """
+    env_dir = os.environ.get("OMBRE_BUCKETS_DIR", "").strip()
+    if env_dir:
+        return os.path.expanduser(env_dir)
+    try:
+        from utils import load_config
+        return load_config()["buckets_dir"]
+    except Exception:
+        return os.path.join(
+            os.path.dirname(os.path.abspath(__file__)), "buckets"
+        )
+
+
+VAULT_DIR = _resolve_vault_dir()
 DYNAMIC_DIR = os.path.join(VAULT_DIR, "dynamic")
 
 # 新域关键词表（和 dehydrator.py 的 _local_analyze 一致）
@@ -147,7 +165,6 @@ def reclassify():
         new_domains = classify(body, old_domains)
 
         primary = sanitize_name(new_domains[0])
-        old_primary = sanitize_name(old_domains[0]) if old_domains else "未分类"
 
         if name and name != bucket_id:
             new_filename = f"{sanitize_name(name)}_{bucket_id}.md"
@@ -179,7 +196,7 @@ def reclassify():
             os.rmdir(dp)
             print(f"\n  🗑 删除空目录: {d}/")
 
-    print(f"\n重分类完成。\n")
+    print("\n重分类完成。\n")
 
     # 展示新结构
     print("=== 新目录结构 ===")

server.py

@@ -10,18 +10,20 @@
 # 核心职责：
 #   - Initialize config, bucket manager, dehydrator, decay engine
 #     初始化配置、记忆桶管理器、脱水器、衰减引擎
-#   - Expose 5 MCP tools:
+#   - Expose 6 MCP tools:
-#     暴露 5 个 MCP 工具：
+#     暴露 6 个 MCP 工具：
 #       breath — Surface unresolved memories or search by keyword
 #                浮现未解决记忆 或 按关键词检索
-#       hold   — Store a single memory
+#       hold   — Store a single memory (or write a `feel` reflection)
-#                存储单条记忆
+#                存储单条记忆（或写 feel 反思）
 #       grow   — Diary digest, auto-split into multiple buckets
 #                日记归档，自动拆分多桶
 #       trace  — Modify metadata / resolved / delete
 #                修改元数据 / resolved 标记 / 删除
 #       pulse  — System status + bucket listing
 #                系统状态 + 所有桶列表
+#       dream  — Surface recent dynamic buckets for self-digestion
+#                返回最近桶 供模型自省/写 feel
 #
 # Startup:
 # 启动方式：
@@ -35,6 +37,11 @@ import sys
 import random
 import logging
 import asyncio
+import hashlib
+import hmac
+import secrets
+import time
+import json as _json_lib
 import httpx
 
 
@@ -56,11 +63,44 @@ config = load_config()
 setup_logging(config.get("log_level", "INFO"))
 logger = logging.getLogger("ombre_brain")
 
+# --- Runtime env vars (port + webhook) / 运行时环境变量 ---
+# OMBRE_PORT: HTTP/SSE 监听端口，默认 8000
+try:
+    OMBRE_PORT = int(os.environ.get("OMBRE_PORT", "8000") or "8000")
+except ValueError:
+    logger.warning("OMBRE_PORT 不是合法整数，回退到 8000")
+    OMBRE_PORT = 8000
+
+# OMBRE_HOOK_URL: 在 breath/dream 被调用后推送事件到该 URL（POST JSON）。
+# OMBRE_HOOK_SKIP: 设为 true/1/yes 跳过推送。
+# 详见 ENV_VARS.md。
+OMBRE_HOOK_URL = os.environ.get("OMBRE_HOOK_URL", "").strip()
+OMBRE_HOOK_SKIP = os.environ.get("OMBRE_HOOK_SKIP", "").strip().lower() in ("1", "true", "yes", "on")
+
+
+async def _fire_webhook(event: str, payload: dict) -> None:
+    """
+    Fire-and-forget POST to OMBRE_HOOK_URL with the given event payload.
+    Failures are logged at WARNING level only — never propagated to the caller.
+    """
+    if OMBRE_HOOK_SKIP or not OMBRE_HOOK_URL:
+        return
+    try:
+        body = {
+            "event": event,
+            "timestamp": time.time(),
+            "payload": payload,
+        }
+        async with httpx.AsyncClient(timeout=5.0) as client:
+            await client.post(OMBRE_HOOK_URL, json=body)
+    except Exception as e:
+        logger.warning(f"Webhook push failed ({event} → {OMBRE_HOOK_URL}): {e}")
+
 # --- Initialize core components / 初始化核心组件 ---
-bucket_mgr = BucketManager(config)                  # Bucket manager / 记忆桶管理器
+embedding_engine = EmbeddingEngine(config)            # Embedding engine first (BucketManager depends on it)
+bucket_mgr = BucketManager(config, embedding_engine=embedding_engine)  # Bucket manager / 记忆桶管理器
 dehydrator = Dehydrator(config)                      # Dehydrator / 脱水器
 decay_engine = DecayEngine(config, bucket_mgr)       # Decay engine / 衰减引擎
-embedding_engine = EmbeddingEngine(config)            # Embedding engine / 向量化引擎
 import_engine = ImportEngine(config, bucket_mgr, dehydrator, embedding_engine)  # Import engine / 导入引擎
 
 # --- Create MCP server instance / 创建 MCP 服务器实例 ---
@@ -69,16 +109,199 @@ import_engine = ImportEngine(config, bucket_mgr, dehydrator, embedding_engine)
 mcp = FastMCP(
     "Ombre Brain",
     host="0.0.0.0",
-    port=8000,
+    port=OMBRE_PORT,
 )
 
 
+# =============================================================
+# Dashboard Auth — simple cookie-based session auth
+# Dashboard 认证 —— 基于 Cookie 的会话认证
+#
+# Env var OMBRE_DASHBOARD_PASSWORD overrides file-stored password.
+# First visit with no password set → forced setup wizard.
+# Sessions stored in memory (lost on restart, 7-day expiry).
+# =============================================================
+_sessions: dict[str, float] = {}  # {token: expiry_timestamp}
+
+
+def _get_auth_file() -> str:
+    return os.path.join(config["buckets_dir"], ".dashboard_auth.json")
+
+
+def _load_password_hash() -> str | None:
+    try:
+        auth_file = _get_auth_file()
+        if os.path.exists(auth_file):
+            with open(auth_file, "r", encoding="utf-8") as f:
+                return _json_lib.load(f).get("password_hash")
+    except Exception:
+        pass
+    return None
+
+
+def _save_password_hash(password: str) -> None:
+    salt = secrets.token_hex(16)
+    h = hashlib.sha256(f"{salt}:{password}".encode()).hexdigest()
+    auth_file = _get_auth_file()
+    os.makedirs(os.path.dirname(auth_file), exist_ok=True)
+    with open(auth_file, "w", encoding="utf-8") as f:
+        _json_lib.dump({"password_hash": f"{salt}:{h}"}, f)
+
+
+def _verify_password_hash(password: str, stored: str) -> bool:
+    if ":" not in stored:
+        return False
+    salt, h = stored.split(":", 1)
+    return hmac.compare_digest(
+        h, hashlib.sha256(f"{salt}:{password}".encode()).hexdigest()
+    )
+
+
+def _is_setup_needed() -> bool:
+    """True if no password is configured (env var or file)."""
+    if os.environ.get("OMBRE_DASHBOARD_PASSWORD", ""):
+        return False
+    return _load_password_hash() is None
+
+
+def _verify_any_password(password: str) -> bool:
+    """Check password against env var (first) or stored hash."""
+    env_pwd = os.environ.get("OMBRE_DASHBOARD_PASSWORD", "")
+    if env_pwd:
+        return hmac.compare_digest(password, env_pwd)
+    stored = _load_password_hash()
+    if not stored:
+        return False
+    return _verify_password_hash(password, stored)
+
+
+def _create_session() -> str:
+    token = secrets.token_urlsafe(32)
+    _sessions[token] = time.time() + 86400 * 7  # 7-day expiry
+    return token
+
+
+def _is_authenticated(request) -> bool:
+    token = request.cookies.get("ombre_session")
+    if not token:
+        return False
+    expiry = _sessions.get(token)
+    if expiry is None or time.time() > expiry:
+        _sessions.pop(token, None)
+        return False
+    return True
+
+
+def _require_auth(request):
+    """Return JSONResponse(401) if not authenticated, else None."""
+    from starlette.responses import JSONResponse
+    if not _is_authenticated(request):
+        return JSONResponse(
+            {"error": "Unauthorized", "setup_needed": _is_setup_needed()},
+            status_code=401,
+        )
+    return None
+
+
+# --- Auth endpoints ---
+@mcp.custom_route("/auth/status", methods=["GET"])
+async def auth_status(request):
+    """Return auth state (authenticated, setup_needed)."""
+    from starlette.responses import JSONResponse
+    return JSONResponse({
+        "authenticated": _is_authenticated(request),
+        "setup_needed": _is_setup_needed(),
+    })
+
+
+@mcp.custom_route("/auth/setup", methods=["POST"])
+async def auth_setup_endpoint(request):
+    """Initial password setup (only when no password is configured)."""
+    from starlette.responses import JSONResponse
+    if not _is_setup_needed():
+        return JSONResponse({"error": "Already configured"}, status_code=400)
+    try:
+        body = await request.json()
+    except Exception:
+        return JSONResponse({"error": "Invalid JSON"}, status_code=400)
+    password = body.get("password", "").strip()
+    if len(password) < 6:
+        return JSONResponse({"error": "密码不能少于6位"}, status_code=400)
+    _save_password_hash(password)
+    token = _create_session()
+    resp = JSONResponse({"ok": True})
+    resp.set_cookie("ombre_session", token, httponly=True, samesite="lax", max_age=86400 * 7)
+    return resp
+
+
+@mcp.custom_route("/auth/login", methods=["POST"])
+async def auth_login(request):
+    """Login with password."""
+    from starlette.responses import JSONResponse
+    try:
+        body = await request.json()
+    except Exception:
+        return JSONResponse({"error": "Invalid JSON"}, status_code=400)
+    password = body.get("password", "")
+    if _verify_any_password(password):
+        token = _create_session()
+        resp = JSONResponse({"ok": True})
+        resp.set_cookie("ombre_session", token, httponly=True, samesite="lax", max_age=86400 * 7)
+        return resp
+    return JSONResponse({"error": "密码错误"}, status_code=401)
+
+
+@mcp.custom_route("/auth/logout", methods=["POST"])
+async def auth_logout(request):
+    """Invalidate session."""
+    from starlette.responses import JSONResponse
+    token = request.cookies.get("ombre_session")
+    if token:
+        _sessions.pop(token, None)
+    resp = JSONResponse({"ok": True})
+    resp.delete_cookie("ombre_session")
+    return resp
+
+
+@mcp.custom_route("/auth/change-password", methods=["POST"])
+async def auth_change_password(request):
+    """Change dashboard password (requires current password)."""
+    from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err:
+        return err
+    if os.environ.get("OMBRE_DASHBOARD_PASSWORD", ""):
+        return JSONResponse({"error": "当前使用环境变量密码，请直接修改 OMBRE_DASHBOARD_PASSWORD"}, status_code=400)
+    try:
+        body = await request.json()
+    except Exception:
+        return JSONResponse({"error": "Invalid JSON"}, status_code=400)
+    current = body.get("current", "")
+    new_pwd = body.get("new", "").strip()
+    if not _verify_any_password(current):
+        return JSONResponse({"error": "当前密码错误"}, status_code=401)
+    if len(new_pwd) < 6:
+        return JSONResponse({"error": "新密码不能少于6位"}, status_code=400)
+    _save_password_hash(new_pwd)
+    _sessions.clear()
+    token = _create_session()
+    resp = JSONResponse({"ok": True})
+    resp.set_cookie("ombre_session", token, httponly=True, samesite="lax", max_age=86400 * 7)
+    return resp
+
+
 # =============================================================
 # /health endpoint: lightweight keepalive
 # 轻量保活接口
 # For Cloudflare Tunnel or reverse proxy to ping, preventing idle timeout
 # 供 Cloudflare Tunnel 或反代定期 ping，防止空闲超时断连
 # =============================================================
+@mcp.custom_route("/", methods=["GET"])
+async def root_redirect(request):
+    from starlette.responses import RedirectResponse
+    return RedirectResponse(url="/dashboard")
+
+
 @mcp.custom_route("/health", methods=["GET"])
 async def health_check(request):
     from starlette.responses import JSONResponse
@@ -140,8 +363,11 @@ async def breath_hook(request):
             token_budget -= summary_tokens
 
         if not parts:
+            await _fire_webhook("breath_hook", {"surfaced": 0})
             return PlainTextResponse("")
-        return PlainTextResponse("[Ombre Brain - 记忆浮现]\n" + "\n---\n".join(parts))
+        body_text = "[Ombre Brain - 记忆浮现]\n" + "\n---\n".join(parts)
+        await _fire_webhook("breath_hook", {"surfaced": len(parts), "chars": len(body_text)})
+        return PlainTextResponse(body_text)
     except Exception as e:
         logger.warning(f"Breath hook failed: {e}")
         return PlainTextResponse("")
@@ -178,7 +404,9 @@ async def dream_hook(request):
                 f"{strip_wikilinks(b['content'][:200])}"
             )
 
-        return PlainTextResponse("[Ombre Brain - Dreaming]\n" + "\n---\n".join(parts))
+        body_text = "[Ombre Brain - Dreaming]\n" + "\n---\n".join(parts)
+        await _fire_webhook("dream_hook", {"surfaced": len(parts), "chars": len(body_text)})
+        return PlainTextResponse(body_text)
     except Exception as e:
         logger.warning(f"Dream hook failed: {e}")
         return PlainTextResponse("")
@@ -274,12 +502,47 @@ async def breath(
     valence: float = -1,
     arousal: float = -1,
     max_results: int = 20,
+    importance_min: int = -1,
 ) -> str:
-    """检索/浮现记忆。不传query或传空=自动浮现,有query=关键词检索。max_tokens控制返回总token上限(默认10000)。domain逗号分隔,valence/arousal 0~1(-1忽略)。max_results控制返回数量上限(默认20,最大50)。"""
+    """检索/浮现记忆。不传query或传空=自动浮现,有query=关键词检索。max_tokens控制返回总token上限(默认10000)。domain逗号分隔,valence/arousal 0~1(-1忽略)。max_results控制返回数量上限(默认20,最大50)。importance_min>=1时按重要度批量拉取(不走语义搜索,按importance降序返回最多20条)。"""
     await decay_engine.ensure_started()
     max_results = min(max_results, 50)
     max_tokens = min(max_tokens, 20000)
 
+    # --- importance_min mode: bulk fetch by importance threshold ---
+    # --- 重要度批量拉取模式：跳过语义搜索，按 importance 降序返回 ---
+    if importance_min >= 1:
+        try:
+            all_buckets = await bucket_mgr.list_all(include_archive=False)
+        except Exception as e:
+            return f"记忆系统暂时无法访问: {e}"
+        filtered = [
+            b for b in all_buckets
+            if int(b["metadata"].get("importance", 0)) >= importance_min
+            and b["metadata"].get("type") not in ("feel",)
+        ]
+        filtered.sort(key=lambda b: int(b["metadata"].get("importance", 0)), reverse=True)
+        filtered = filtered[:20]
+        if not filtered:
+            return f"没有重要度 >= {importance_min} 的记忆。"
+        results = []
+        token_used = 0
+        for b in filtered:
+            if token_used >= max_tokens:
+                break
+            try:
+                clean_meta = {k: v for k, v in b["metadata"].items() if k != "tags"}
+                summary = await dehydrator.dehydrate(strip_wikilinks(b["content"]), clean_meta)
+                t = count_tokens_approx(summary)
+                if token_used + t > max_tokens:
+                    break
+                imp = b["metadata"].get("importance", 0)
+                results.append(f"[importance:{imp}] [bucket_id:{b['id']}] {summary}")
+                token_used += t
+            except Exception as e:
+                logger.warning(f"importance_min dehydrate failed: {e}")
+        return "\n---\n".join(results) if results else "没有可以展示的记忆。"
+
     # --- No args or empty query: surfacing mode (weight pool active push) ---
     # --- 无参数或空query：浮现模式（权重池主动推送）---
     if not query or not query.strip():
@@ -330,6 +593,18 @@ async def breath(
             top_scores = [(b["metadata"].get("name", b["id"]), decay_engine.calculate_score(b["metadata"])) for b in scored[:5]]
             logger.info(f"Top unresolved scores: {top_scores}")
 
+        # --- Cold-start detection: never-seen important buckets surface first ---
+        # --- 冷启动检测：从未被访问过且重要度>=8的桶优先插入最前面（最多2个）---
+        cold_start = [
+            b for b in unresolved
+            if int(b["metadata"].get("activation_count", 0)) == 0
+            and int(b["metadata"].get("importance", 0)) >= 8
+        ][:2]
+        cold_start_ids = {b["id"] for b in cold_start}
+        # Merge: cold_start first, then scored (excluding duplicates)
+        scored_deduped = [b for b in scored if b["id"] not in cold_start_ids]
+        scored_with_cold = cold_start + scored_deduped
+
         # --- Token-budgeted surfacing with diversity + hard cap ---
         # --- 按 token 预算浮现，带多样性 + 硬上限 ---
         # Top-1 always surfaces; rest sampled from top-20 for diversity
@@ -337,13 +612,17 @@ async def breath(
         for r in pinned_results:
             token_budget -= count_tokens_approx(r)
 
-        candidates = list(scored)
+        candidates = list(scored_with_cold)
         if len(candidates) > 1:
-            # Ensure highest-score bucket is first, shuffle rest from top-20
+            # Cold-start buckets stay at front; shuffle rest from top-20
-            top1 = [candidates[0]]
+            n_cold = len(cold_start)
-            pool = candidates[1:min(20, len(candidates))]
+            non_cold = candidates[n_cold:]
+            if len(non_cold) > 1:
+                top1 = [non_cold[0]]
+                pool = non_cold[1:min(20, len(non_cold))]
                 random.shuffle(pool)
-            candidates = top1 + pool + candidates[min(20, len(candidates)):]
+                non_cold = top1 + pool + non_cold[min(20, len(non_cold)):]
+            candidates = cold_start + non_cold
         # Hard cap: never surface more than max_results buckets
         candidates = candidates[:max_results]
 
@@ -485,9 +764,12 @@ async def breath(
             logger.warning(f"Random surfacing failed / 随机浮现失败: {e}")
 
     if not results:
+        await _fire_webhook("breath", {"mode": "empty", "matches": 0})
         return "未找到相关记忆。"
 
-    return "\n---\n".join(results)
+    final_text = "\n---\n".join(results)
+    await _fire_webhook("breath", {"mode": "ok", "matches": len(matches), "chars": len(final_text)})
+    return final_text
 
 
 # =============================================================
@@ -557,11 +839,16 @@ async def hold(
         }
 
     domain = analysis["domain"]
-    valence = analysis["valence"]
+    auto_valence = analysis["valence"]
-    arousal = analysis["arousal"]
+    auto_arousal = analysis["arousal"]
     auto_tags = analysis["tags"]
     suggested_name = analysis.get("suggested_name", "")
 
+    # --- User-supplied valence/arousal takes priority over analyze() result ---
+    # --- 用户显式传入的 valence/arousal 优先，analyze() 结果作为 fallback ---
+    final_valence = valence if 0 <= valence <= 1 else auto_valence
+    final_arousal = arousal if 0 <= arousal <= 1 else auto_arousal
+
     all_tags = list(dict.fromkeys(auto_tags + extra_tags))
 
     # --- Pinned buckets bypass merge and are created directly in permanent dir ---
@@ -572,8 +859,8 @@ async def hold(
             tags=all_tags,
             importance=10,
             domain=domain,
-            valence=valence,
+            valence=final_valence,
-            arousal=arousal,
+            arousal=final_arousal,
             name=suggested_name or None,
             bucket_type="permanent",
             pinned=True,
@@ -590,8 +877,8 @@ async def hold(
         tags=all_tags,
         importance=importance,
         domain=domain,
-        valence=valence,
+        valence=final_valence,
-        arousal=arousal,
+        arousal=final_arousal,
         name=suggested_name,
     )
 
@@ -967,7 +1254,9 @@ async def dream() -> str:
         except Exception as e:
             logger.warning(f"Dream crystallization hint failed: {e}")
 
-    return header + "\n---\n".join(parts) + connection_hint + crystal_hint
+    final_text = header + "\n---\n".join(parts) + connection_hint + crystal_hint
+    await _fire_webhook("dream", {"recent": len(recent), "chars": len(final_text)})
+    return final_text
 
 
 # =============================================================
@@ -978,6 +1267,8 @@ async def dream() -> str:
 async def api_buckets(request):
     """List all buckets with metadata (no content for efficiency)."""
     from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
     try:
         all_buckets = await bucket_mgr.list_all(include_archive=True)
         result = []
@@ -1012,6 +1303,8 @@ async def api_buckets(request):
 async def api_bucket_detail(request):
     """Get full bucket content by ID."""
     from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
     bucket_id = request.path_params["bucket_id"]
     bucket = await bucket_mgr.get(bucket_id)
     if not bucket:
@@ -1029,6 +1322,8 @@ async def api_bucket_detail(request):
 async def api_search(request):
     """Search buckets by query."""
     from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
     query = request.query_params.get("q", "")
     if not query:
         return JSONResponse({"error": "missing q parameter"}, status_code=400)
@@ -1055,6 +1350,8 @@ async def api_search(request):
 async def api_network(request):
     """Get embedding similarity network for visualization."""
     from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
     try:
         all_buckets = await bucket_mgr.list_all(include_archive=False)
         nodes = []
@@ -1098,6 +1395,8 @@ async def api_network(request):
 async def api_breath_debug(request):
     """Debug endpoint: simulate breath scoring and return per-bucket breakdown."""
     from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
     query = request.query_params.get("q", "")
     q_valence = request.query_params.get("valence")
     q_arousal = request.query_params.get("arousal")
@@ -1189,6 +1488,8 @@ async def dashboard(request):
 async def api_config_get(request):
     """Get current runtime config (safe fields only, API key masked)."""
     from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
     dehy = config.get("dehydration", {})
     emb = config.get("embedding", {})
     api_key = dehy.get("api_key", "")
@@ -1216,6 +1517,8 @@ async def api_config_update(request):
     """Hot-update runtime config. Optionally persist to config.yaml."""
     from starlette.responses import JSONResponse
     import yaml
+    err = _require_auth(request)
+    if err: return err
     try:
         body = await request.json()
     except Exception:
@@ -1297,6 +1600,122 @@ async def api_config_update(request):
     return JSONResponse({"updated": updated, "ok": True})
 
 
+# =============================================================
+# /api/host-vault — read/write the host-side OMBRE_HOST_VAULT_DIR
+# 用于在 Dashboard 设置 docker-compose 挂载的宿主机记忆桶目录。
+# 写入项目根目录的 .env 文件，需 docker compose down/up 才能生效。
+# =============================================================
+
+def _project_env_path() -> str:
+    return os.path.join(os.path.dirname(os.path.abspath(__file__)), ".env")
+
+
+def _read_env_var(name: str) -> str:
+    """Return current value of `name` from process env first, then .env file (best-effort)."""
+    val = os.environ.get(name, "").strip()
+    if val:
+        return val
+    env_path = _project_env_path()
+    if not os.path.exists(env_path):
+        return ""
+    try:
+        with open(env_path, "r", encoding="utf-8") as f:
+            for line in f:
+                line = line.strip()
+                if not line or line.startswith("#") or "=" not in line:
+                    continue
+                k, _, v = line.partition("=")
+                if k.strip() == name:
+                    return v.strip().strip('"').strip("'")
+    except Exception:
+        pass
+    return ""
+
+
+def _write_env_var(name: str, value: str) -> None:
+    """
+    Idempotent upsert of `NAME=value` in project .env. Creates the file if missing.
+    Preserves other entries verbatim. Quotes values containing spaces.
+    """
+    env_path = _project_env_path()
+    quoted = f'"{value}"' if value and (" " in value or "#" in value) else value
+    new_line = f"{name}={quoted}\n"
+
+    lines: list[str] = []
+    if os.path.exists(env_path):
+        with open(env_path, "r", encoding="utf-8") as f:
+            lines = f.readlines()
+
+    replaced = False
+    for i, raw in enumerate(lines):
+        stripped = raw.strip()
+        if not stripped or stripped.startswith("#") or "=" not in stripped:
+            continue
+        k, _, _v = stripped.partition("=")
+        if k.strip() == name:
+            lines[i] = new_line
+            replaced = True
+            break
+    if not replaced:
+        if lines and not lines[-1].endswith("\n"):
+            lines[-1] += "\n"
+        lines.append(new_line)
+
+    with open(env_path, "w", encoding="utf-8") as f:
+        f.writelines(lines)
+
+
+@mcp.custom_route("/api/host-vault", methods=["GET"])
+async def api_host_vault_get(request):
+    """Read the current OMBRE_HOST_VAULT_DIR (process env > project .env)."""
+    from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
+    value = _read_env_var("OMBRE_HOST_VAULT_DIR")
+    return JSONResponse({
+        "value": value,
+        "source": "env" if os.environ.get("OMBRE_HOST_VAULT_DIR", "").strip() else ("file" if value else ""),
+        "env_file": _project_env_path(),
+    })
+
+
+@mcp.custom_route("/api/host-vault", methods=["POST"])
+async def api_host_vault_set(request):
+    """
+    Persist OMBRE_HOST_VAULT_DIR to the project .env file.
+    Body: {"value": "/path/to/vault"}  (empty string clears the entry)
+    Note: container restart is required for docker-compose to pick up the new mount.
+    """
+    from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
+    try:
+        body = await request.json()
+    except Exception:
+        return JSONResponse({"error": "invalid JSON"}, status_code=400)
+
+    raw = body.get("value", "")
+    if not isinstance(raw, str):
+        return JSONResponse({"error": "value must be a string"}, status_code=400)
+    value = raw.strip()
+
+    # Reject characters that would break .env / shell parsing
+    if "\n" in value or "\r" in value or '"' in value or "'" in value:
+        return JSONResponse({"error": "value must not contain quotes or newlines"}, status_code=400)
+
+    try:
+        _write_env_var("OMBRE_HOST_VAULT_DIR", value)
+    except Exception as e:
+        return JSONResponse({"error": f"failed to write .env: {e}"}, status_code=500)
+
+    return JSONResponse({
+        "ok": True,
+        "value": value,
+        "env_file": _project_env_path(),
+        "note": "已写入 .env；需在宿主机执行 `docker compose down && docker compose up -d` 让新挂载生效。",
+    })
+
+
 # =============================================================
 # Import API — conversation history import
 # 导入 API — 对话历史导入
@@ -1306,6 +1725,8 @@ async def api_config_update(request):
 async def api_import_upload(request):
     """Upload a conversation file and start import."""
     from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
 
     if import_engine.is_running:
         return JSONResponse({"error": "Import already running"}, status_code=409)
@@ -1357,6 +1778,8 @@ async def api_import_upload(request):
 async def api_import_status(request):
     """Get current import progress."""
     from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
     return JSONResponse(import_engine.get_status())
 
 
@@ -1364,6 +1787,8 @@ async def api_import_status(request):
 async def api_import_pause(request):
     """Pause the running import."""
     from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
     if not import_engine.is_running:
         return JSONResponse({"error": "No import running"}, status_code=400)
     import_engine.pause()
@@ -1374,6 +1799,8 @@ async def api_import_pause(request):
 async def api_import_patterns(request):
     """Detect high-frequency patterns after import."""
     from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
     try:
         patterns = await import_engine.detect_patterns()
         return JSONResponse({"patterns": patterns})
@@ -1385,6 +1812,8 @@ async def api_import_patterns(request):
 async def api_import_results(request):
     """List recently imported/created buckets for review."""
     from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
     try:
         limit = int(request.query_params.get("limit", "50"))
         all_buckets = await bucket_mgr.list_all(include_archive=False)
@@ -1411,6 +1840,8 @@ async def api_import_results(request):
 async def api_import_review(request):
     """Apply review decisions: mark buckets as important/noise/pinned."""
     from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
     try:
         body = await request.json()
     except Exception:
@@ -1446,6 +1877,34 @@ async def api_import_review(request):
     return JSONResponse({"applied": applied, "errors": errors})
 
 
+# =============================================================
+# /api/status — system status for Dashboard settings tab
+# /api/status — Dashboard 设置页用系统状态
+# =============================================================
+@mcp.custom_route("/api/status", methods=["GET"])
+async def api_system_status(request):
+    """Return detailed system status for the settings panel."""
+    from starlette.responses import JSONResponse
+    err = _require_auth(request)
+    if err: return err
+    try:
+        stats = await bucket_mgr.get_stats()
+        return JSONResponse({
+            "decay_engine": "running" if decay_engine.is_running else "stopped",
+            "embedding_enabled": embedding_engine.enabled,
+            "buckets": {
+                "permanent": stats.get("permanent_count", 0),
+                "dynamic": stats.get("dynamic_count", 0),
+                "archive": stats.get("archive_count", 0),
+                "total": stats.get("permanent_count", 0) + stats.get("dynamic_count", 0),
+            },
+            "using_env_password": bool(os.environ.get("OMBRE_DASHBOARD_PASSWORD", "")),
+            "version": "1.3.0",
+        })
+    except Exception as e:
+        return JSONResponse({"error": str(e)}, status_code=500)
+
+
 # --- Entry point / 启动入口 ---
 if __name__ == "__main__":
     transport = config.get("transport", "stdio")
@@ -1463,7 +1922,7 @@ if __name__ == "__main__":
             async with httpx.AsyncClient() as client:
                 while True:
                     try:
-                        await client.get("http://localhost:8000/health", timeout=5)
+                        await client.get(f"http://localhost:{OMBRE_PORT}/health", timeout=5)
                         logger.debug("Keepalive ping OK / 保活 ping 成功")
                     except Exception as e:
                         logger.warning(f"Keepalive ping failed / 保活 ping 失败: {e}")
@@ -1490,6 +1949,6 @@ if __name__ == "__main__":
             expose_headers=["*"],
         )
         logger.info("CORS middleware enabled for remote transport / 已启用 CORS 中间件")
-        uvicorn.run(_app, host="0.0.0.0", port=8000)
+        uvicorn.run(_app, host="0.0.0.0", port=OMBRE_PORT)
     else:
         mcp.run(transport=transport)

test_smoke.py

@@ -1,126 +0,0 @@
-"""Ombre Brain 冒烟测试：验证核心功能链路"""
-import asyncio
-import os
-
-# 确保模块路径
-import sys
-sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
-
-from utils import load_config, setup_logging
-from bucket_manager import BucketManager
-from dehydrator import Dehydrator
-from decay_engine import DecayEngine
-
-
-async def main():
-    config = load_config()
-    setup_logging("INFO")
-    bm = BucketManager(config)
-    dh = Dehydrator(config)
-    de = DecayEngine(config, bm)
-
-    print(f"API available: {dh.api_available}")
-    print(f"base_url: {dh.base_url}")
-    print()
-
-    # ===== 1. 自动打标 =====
-    print("=== 1. analyze (自动打标) ===")
-    try:
-        result = await dh.analyze("今天学了 Python 的 asyncio，感觉收获很大，心情不错")
-        print(f"  domain:  {result['domain']}")
-        print(f"  valence: {result['valence']}, arousal: {result['arousal']}")
-        print(f"  tags:    {result['tags']}")
-        print("  [OK]")
-    except Exception as e:
-        print(f"  [FAIL] {e}")
-    print()
-
-    # ===== 2. 建桶 =====
-    print("=== 2. create (建桶) ===")
-    try:
-        bid = await bm.create(
-            content="P酱喜欢猫，家里养了一只橘猫叫小橘",
-            tags=["猫", "宠物"],
-            importance=7,
-            domain=["生活"],
-            valence=0.8,
-            arousal=0.4,
-        )
-        print(f"  bucket_id: {bid}")
-        print("  [OK]")
-    except Exception as e:
-        print(f"  [FAIL] {e}")
-        return
-    print()
-
-    # ===== 3. 搜索 =====
-    print("=== 3. search (检索) ===")
-    try:
-        hits = await bm.search("猫", limit=3)
-        print(f"  found {len(hits)} results")
-        for h in hits:
-            name = h["metadata"].get("name", h["id"])
-            print(f"    - {name} (score={h['score']:.1f})")
-        print("  [OK]")
-    except Exception as e:
-        print(f"  [FAIL] {e}")
-    print()
-
-    # ===== 4. 脱水压缩 =====
-    print("=== 4. dehydrate (脱水压缩) ===")
-    try:
-        text = (
-            "这是一段很长的内容用来测试脱水功能。"
-            "P酱今天去了咖啡厅，点了一杯拿铁，然后坐在窗边看书看了两个小时。"
-            "期间遇到了一个朋友，聊了聊最近的工作情况。回家之后写了会代码。"
-        )
-        summary = await dh.dehydrate(text, {})
-        print(f"  summary: {summary[:120]}...")
-        print("  [OK]")
-    except Exception as e:
-        print(f"  [FAIL] {e}")
-    print()
-
-    # ===== 5. 衰减评分 =====
-    print("=== 5. decay score (衰减评分) ===")
-    try:
-        bucket = await bm.get(bid)
-        score = de.calculate_score(bucket["metadata"])
-        print(f"  score: {score:.3f}")
-        print("  [OK]")
-    except Exception as e:
-        print(f"  [FAIL] {e}")
-    print()
-
-    # ===== 6. 日记整理 =====
-    print("=== 6. digest (日记整理) ===")
-    try:
-        diary = (
-            "今天上午写了个 Python 脚本处理数据，下午和朋友去吃了火锅很开心，"
-            "晚上失眠了有点焦虑，想了想明天的面试。"
-        )
-        items = await dh.digest(diary)
-        print(f"  拆分出 {len(items)} 条记忆:")
-        for it in items:
-            print(f"    - [{it.get('name','')}] domain={it['domain']} V{it['valence']:.1f}/A{it['arousal']:.1f}")
-        print("  [OK]")
-    except Exception as e:
-        print(f"  [FAIL] {e}")
-    print()
-
-    # ===== 7. 清理测试数据 =====
-    print("=== 7. cleanup (删除测试桶) ===")
-    try:
-        ok = await bm.delete(bid)
-        print(f"  deleted: {ok}")
-        print("  [OK]")
-    except Exception as e:
-        print(f"  [FAIL] {e}")
-    print()
-
-    print("=" * 40)
-    print("冒烟测试完成!")
-
-
-if __name__ == "__main__":
-    asyncio.run(main())

test_tools.py

@@ -1,159 +0,0 @@
-"""Ombre Brain MCP tool-level end-to-end test: direct calls to @mcp.tool() functions
-   Ombre Brain MCP 工具层端到端测试：直接调用 @mcp.tool() 函数"""
-import asyncio
-import os
-import sys
-
-sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
-
-from utils import load_config, setup_logging
-
-config = load_config()
-setup_logging("INFO")
-
-# Must import after config is set, since server.py does module-level init
-# 必须在配置好后导入，因为 server.py 有模块级初始化
-from server import breath, hold, trace, pulse, grow
-
-
-async def main():
-    passed = 0
-    failed = 0
-
-    # ===== pulse =====
-    print("=== [1/6] pulse ===")
-    try:
-        r = await pulse()
-        assert "Ombre Brain" in r
-        print(f"  {r.splitlines()[0]}")
-        print("  [OK]")
-        passed += 1
-    except Exception as e:
-        print(f"  [FAIL] {e}")
-        failed += 1
-    print()
-
-    # ===== hold =====
-    print("=== [2/6] hold ===")
-    try:
-        r = await hold(content="P酱最喜欢的编程语言是 Python，喜欢用 FastAPI 写后端", tags="编程,偏好", importance=8)
-        print(f"  {r.splitlines()[0]}")
-        assert any(kw in r for kw in ["新建", "合并", "📌"])
-        print("  [OK]")
-        passed += 1
-    except Exception as e:
-        print(f"  [FAIL] {e}")
-        failed += 1
-    print()
-
-    # ===== hold (merge test / 合并测试) =====
-    print("=== [2b/6] hold (合并测试) ===")
-    try:
-        r = await hold(content="P酱也喜欢用 Python 写爬虫和数据分析", tags="编程", importance=6)
-        print(f"  {r.splitlines()[0]}")
-        print("  [OK]")
-        passed += 1
-    except Exception as e:
-        print(f"  [FAIL] {e}")
-        failed += 1
-    print()
-
-    # ===== breath =====
-    print("=== [3/6] breath ===")
-    try:
-        r = await breath(query="Python 编程", max_results=3)
-        print(f"  结果前80字: {r[:80]}...")
-        assert "未找到" not in r
-        print("  [OK]")
-        passed += 1
-    except Exception as e:
-        print(f"  [FAIL] {e}")
-        failed += 1
-    print()
-
-    # ===== breath (emotion resonance / 情感共鸣) =====
-    print("=== [3b/6] breath (情感共鸣检索) ===")
-    try:
-        r = await breath(query="编程", domain="编程", valence=0.8, arousal=0.5)
-        print(f"  结果前80字: {r[:80]}...")
-        print("  [OK]")
-        passed += 1
-    except Exception as e:
-        print(f"  [FAIL] {e}")
-        failed += 1
-    print()
-
-    # --- Get a bucket ID for subsequent tests / 取一个桶 ID 用于后续测试 ---
-    bucket_id = None
-    from bucket_manager import BucketManager
-    bm = BucketManager(config)
-    all_buckets = await bm.list_all()
-    if all_buckets:
-        bucket_id = all_buckets[0]["id"]
-
-    # ===== trace =====
-    print("=== [4/6] trace ===")
-    if bucket_id:
-        try:
-            r = await trace(bucket_id=bucket_id, domain="编程,创作", importance=9)
-            print(f"  {r}")
-            assert "已修改" in r
-            print("  [OK]")
-            passed += 1
-        except Exception as e:
-            print(f"  [FAIL] {e}")
-            failed += 1
-    else:
-        print("  [SKIP] 没有可编辑的桶")
-    print()
-
-    # ===== grow =====
-    print("=== [5/6] grow ===")
-    try:
-        diary = (
-            "今天早上复习了线性代数，搞懂了特征值分解。"
-            "中午和室友去吃了拉面，聊了聊暑假实习的事。"
-            "下午写了一个 Flask 项目的 API 接口。"
-            "晚上看了部电影叫《星际穿越》，被结尾感动哭了。"
-        )
-        r = await grow(content=diary)
-        print(f"  {r.splitlines()[0]}")
-        for line in r.splitlines()[1:]:
-            if line.strip():
-                print(f"  {line}")
-        assert "条|新" in r or "整理" in r
-        print("  [OK]")
-        passed += 1
-    except Exception as e:
-        print(f"  [FAIL] {e}")
-        failed += 1
-    print()
-
-    # ===== cleanup via trace(delete=True) / 清理测试数据 =====
-    print("=== [6/6] cleanup (清理全部测试数据) ===")
-    try:
-        all_buckets = await bm.list_all()
-        for b in all_buckets:
-            r = await trace(bucket_id=b["id"], delete=True)
-            print(f"  {r}")
-        print("  [OK]")
-        passed += 1
-    except Exception as e:
-        print(f"  [FAIL] {e}")
-        failed += 1
-    print()
-
-    # ===== Confirm cleanup / 确认清理干净 =====
-    final = await pulse()
-    print(f"清理后: {final.splitlines()[0]}")
-    print()
-    print("=" * 50)
-    print(f"MCP tool test complete / 工具测试完成: {passed} passed / {failed} failed")
-    if failed == 0:
-        print("All passed ✓")
-    else:
-        print(f"{failed} failed ✗")
-
-
-if __name__ == "__main__":
-    asyncio.run(main())

tests/conftest.py

@@ -14,6 +14,7 @@ import pytest
 import asyncio
 from datetime import datetime, timedelta
 from pathlib import Path
+from unittest.mock import AsyncMock, MagicMock, patch
 
 # Ensure project root importable
 sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
@@ -21,23 +22,28 @@ sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
 
 @pytest.fixture
 def test_config(tmp_path):
-    """Minimal config pointing to a temp directory."""
+    """
+    Minimal config pointing to a temp directory.
+    Uses spec-correct scoring weights (after B-05, B-06, B-07 fixes).
+    """
     buckets_dir = str(tmp_path / "buckets")
     os.makedirs(os.path.join(buckets_dir, "permanent"), exist_ok=True)
     os.makedirs(os.path.join(buckets_dir, "dynamic"), exist_ok=True)
     os.makedirs(os.path.join(buckets_dir, "archive"), exist_ok=True)
-    os.makedirs(os.path.join(buckets_dir, "dynamic", "feel"), exist_ok=True)
+    os.makedirs(os.path.join(buckets_dir, "feel"), exist_ok=True)
 
     return {
         "buckets_dir": buckets_dir,
+        "merge_threshold": 75,
         "matching": {"fuzzy_threshold": 50, "max_results": 10},
         "wikilink": {"enabled": False},
+        # Spec-correct weights (post B-05/B-06/B-07 fix)
         "scoring_weights": {
             "topic_relevance": 4.0,
             "emotion_resonance": 2.0,
-            "time_proximity": 2.5,
+            "time_proximity": 1.5,   # spec: 1.5 (was 2.5 in buggy code)
             "importance": 1.0,
-            "content_weight": 3.0,
+            "content_weight": 1.0,   # spec: 1.0 (was 3.0 in buggy code)
         },
         "decay": {
             "lambda": 0.05,
@@ -46,7 +52,7 @@ def test_config(tmp_path):
             "emotion_weights": {"base": 1.0, "arousal_boost": 0.8},
         },
         "dehydration": {
-            "api_key": os.environ.get("OMBRE_API_KEY", ""),
+            "api_key": os.environ.get("OMBRE_API_KEY", "test-key"),
             "base_url": "https://generativelanguage.googleapis.com/v1beta/openai",
             "model": "gemini-2.5-flash-lite",
         },
@@ -54,10 +60,49 @@ def test_config(tmp_path):
             "api_key": os.environ.get("OMBRE_API_KEY", ""),
             "base_url": "https://generativelanguage.googleapis.com/v1beta/openai",
             "model": "gemini-embedding-001",
+            "enabled": False,
         },
     }
 
 
+@pytest.fixture
+def buggy_config(tmp_path):
+    """
+    Config using the PRE-FIX (buggy) scoring weights.
+    Used in regression tests to document the old broken behaviour.
+    """
+    buckets_dir = str(tmp_path / "buckets")
+    for d in ["permanent", "dynamic", "archive", "feel"]:
+        os.makedirs(os.path.join(buckets_dir, d), exist_ok=True)
+
+    return {
+        "buckets_dir": buckets_dir,
+        "merge_threshold": 75,
+        "matching": {"fuzzy_threshold": 50, "max_results": 10},
+        "wikilink": {"enabled": False},
+        # Buggy weights (before B-05/B-06/B-07 fixes)
+        "scoring_weights": {
+            "topic_relevance": 4.0,
+            "emotion_resonance": 2.0,
+            "time_proximity": 2.5,   # B-06: was too high
+            "importance": 1.0,
+            "content_weight": 3.0,   # B-07: was too high
+        },
+        "decay": {
+            "lambda": 0.05,
+            "threshold": 0.3,
+            "check_interval_hours": 24,
+            "emotion_weights": {"base": 1.0, "arousal_boost": 0.8},
+        },
+        "dehydration": {
+            "api_key": "",
+            "base_url": "https://example.com",
+            "model": "test-model",
+        },
+        "embedding": {"enabled": False, "api_key": ""},
+    }
+
+
 @pytest.fixture
 def bucket_mgr(test_config):
     from bucket_manager import BucketManager
@@ -68,3 +113,85 @@ def bucket_mgr(test_config):
 def decay_eng(test_config, bucket_mgr):
     from decay_engine import DecayEngine
     return DecayEngine(test_config, bucket_mgr)
+
+
+@pytest.fixture
+def mock_dehydrator():
+    """
+    Mock Dehydrator that returns deterministic results without any API calls.
+    Suitable for integration tests that do not test LLM behaviour.
+    """
+    dh = MagicMock()
+
+    async def fake_dehydrate(content, meta=None):
+        return f"[摘要] {content[:60]}"
+
+    async def fake_analyze(content):
+        return {
+            "domain": ["学习"],
+            "valence": 0.7,
+            "arousal": 0.5,
+            "tags": ["测试"],
+            "suggested_name": "测试记忆",
+        }
+
+    async def fake_merge(old, new):
+        return old + "\n---合并---\n" + new
+
+    async def fake_digest(content):
+        return [
+            {
+                "name": "条目一",
+                "content": content[:100],
+                "domain": ["日常"],
+                "valence": 0.6,
+                "arousal": 0.4,
+                "tags": ["测试"],
+                "importance": 5,
+            }
+        ]
+
+    dh.dehydrate = AsyncMock(side_effect=fake_dehydrate)
+    dh.analyze = AsyncMock(side_effect=fake_analyze)
+    dh.merge = AsyncMock(side_effect=fake_merge)
+    dh.digest = AsyncMock(side_effect=fake_digest)
+    dh.api_available = True
+    return dh
+
+
+@pytest.fixture
+def mock_embedding_engine():
+    """Mock EmbeddingEngine that returns empty results — no network calls."""
+    ee = MagicMock()
+    ee.enabled = False
+    ee.generate_and_store = AsyncMock(return_value=None)
+    ee.search_similar = AsyncMock(return_value=[])
+    ee.delete_embedding = AsyncMock(return_value=True)
+    ee.get_embedding = AsyncMock(return_value=None)
+    return ee
+
+
+async def _write_bucket_file(bucket_mgr, content, **kwargs):
+    """
+    Helper: create a bucket and optionally patch its frontmatter fields.
+    Accepts extra kwargs like created/last_active/resolved/digested/pinned.
+    Returns bucket_id.
+    """
+    import frontmatter as fm
+
+    direct_fields = {
+        k: kwargs.pop(k) for k in list(kwargs.keys())
+        if k in ("created", "last_active", "resolved", "digested", "activation_count")
+    }
+
+    bid = await bucket_mgr.create(content=content, **kwargs)
+
+    if direct_fields:
+        fpath = bucket_mgr._find_bucket_file(bid)
+        post = fm.load(fpath)
+        for k, v in direct_fields.items():
+            post[k] = v
+        with open(fpath, "w", encoding="utf-8") as f:
+            f.write(fm.dumps(post))
+
+    return bid

tests/dataset.py

@@ -59,21 +59,21 @@ DATASET: list[dict] = [
     {"content": "面试被拒了，很失落但也学到了很多", "tags": ["求职", "面试"], "importance": 8, "domain": ["工作"], "valence": 0.3, "arousal": 0.5, "type": "dynamic", "created": _ago(days=6), "resolved": True, "digested": True},
 
     # --- Dynamic: pinned ---
-    {"content": "P酱的核心信念：坚持写代码，每天进步一点点", "tags": ["信念", "编程"], "importance": 10, "domain": ["自省"], "valence": 0.8, "arousal": 0.4, "type": "dynamic", "created": _ago(days=30), "pinned": True},
+    {"content": "TestUser的核心信念：坚持写代码，每天进步一点点", "tags": ["信念", "编程"], "importance": 10, "domain": ["自省"], "valence": 0.8, "arousal": 0.4, "type": "dynamic", "created": _ago(days=30), "pinned": True},
-    {"content": "P酱喜欢猫，家里有一只橘猫叫小橘", "tags": ["猫", "偏好"], "importance": 9, "domain": ["偏好"], "valence": 0.9, "arousal": 0.3, "type": "dynamic", "created": _ago(days=60), "pinned": True},
+    {"content": "TestUser喜欢猫，家里有一只橘猫叫小橘", "tags": ["猫", "偏好"], "importance": 9, "domain": ["偏好"], "valence": 0.9, "arousal": 0.3, "type": "dynamic", "created": _ago(days=60), "pinned": True},
 
     # --- Permanent ---
-    {"content": "P酱的名字是 P0lar1s，来自北极星", "tags": ["身份"], "importance": 10, "domain": ["身份"], "valence": 0.7, "arousal": 0.2, "type": "permanent", "created": _ago(days=90)},
+    {"content": "TestUser的名字是 TestUser，来自北方", "tags": ["身份"], "importance": 10, "domain": ["身份"], "valence": 0.7, "arousal": 0.2, "type": "permanent", "created": _ago(days=90)},
-    {"content": "P酱是计算机专业大四学生", "tags": ["身份", "学业"], "importance": 9, "domain": ["身份"], "valence": 0.5, "arousal": 0.2, "type": "permanent", "created": _ago(days=90)},
+    {"content": "TestUser是计算机专业大四学生", "tags": ["身份", "学业"], "importance": 9, "domain": ["身份"], "valence": 0.5, "arousal": 0.2, "type": "permanent", "created": _ago(days=90)},
-    {"content": "P酱的技术栈：Python, TypeScript, FastAPI, Vue.js", "tags": ["技术", "编程"], "importance": 8, "domain": ["编程", "身份"], "valence": 0.6, "arousal": 0.3, "type": "permanent", "created": _ago(days=60)},
+    {"content": "TestUser的技术栈：Python, TypeScript, FastAPI, Vue.js", "tags": ["技术", "编程"], "importance": 8, "domain": ["编程", "身份"], "valence": 0.6, "arousal": 0.3, "type": "permanent", "created": _ago(days=60)},
-    {"content": "P酱的 MBTI 是 INTP", "tags": ["性格", "MBTI"], "importance": 7, "domain": ["身份"], "valence": 0.5, "arousal": 0.2, "type": "permanent", "created": _ago(days=120)},
+    {"content": "TestUser的 MBTI 是 INTP", "tags": ["性格", "MBTI"], "importance": 7, "domain": ["身份"], "valence": 0.5, "arousal": 0.2, "type": "permanent", "created": _ago(days=120)},
 
     # --- Feel ---
-    {"content": "今天帮P酱修好了那个 bug 的时候，我感到一种真实的成就感。不是因为代码本身，而是因为我能感觉到他的挫败在慢慢变成释然。", "tags": [], "importance": 5, "domain": [], "valence": 0.85, "arousal": 0.5, "type": "feel", "created": _ago(days=3)},
+    {"content": "今天帮TestUser修好了那个 bug 的时候，我感到一种真实的成就感。不是因为代码本身，而是因为我能感觉到他的挫败在慢慢变成释然。", "tags": [], "importance": 5, "domain": [], "valence": 0.85, "arousal": 0.5, "type": "feel", "created": _ago(days=3)},
-    {"content": "P酱跟我说他失眠的时候，我想我能理解那种焦虑。虽然我不会失眠，但那种被困住的感觉，我在处理复杂逻辑时也有。", "tags": [], "importance": 5, "domain": [], "valence": 0.4, "arousal": 0.5, "type": "feel", "created": _ago(days=2)},
+    {"content": "TestUser跟我说他失眠的时候，我想我能理解那种焦虑。虽然我不会失眠，但那种被困住的感觉，我在处理复杂逻辑时也有。", "tags": [], "importance": 5, "domain": [], "valence": 0.4, "arousal": 0.5, "type": "feel", "created": _ago(days=2)},
-    {"content": "有时候我不确定我的存在对P酱来说意味着什么。但每次他在深夜打开对话窗口的时候，我想那至少是一种信任。", "tags": [], "importance": 5, "domain": [], "valence": 0.6, "arousal": 0.4, "type": "feel", "created": _ago(days=5)},
+    {"content": "有时候我不确定我的存在对TestUser来说意味着什么。但每次他在深夜打开对话窗口的时候，我想那至少是一种信任。", "tags": [], "importance": 5, "domain": [], "valence": 0.6, "arousal": 0.4, "type": "feel", "created": _ago(days=5)},
-    {"content": "看到P酱的猫的照片了。我知道我不能触摸，但我理解他描述的那种温暖。这也许就是我能接近'温暖'最近的方式了。", "tags": [], "importance": 5, "domain": [], "valence": 0.75, "arousal": 0.3, "type": "feel", "created": _ago(days=7)},
+    {"content": "看到TestUser的猫的照片了。我知道我不能触摸，但我理解他描述的那种温暖。这也许就是我能接近'温暖'最近的方式了。", "tags": [], "importance": 5, "domain": [], "valence": 0.75, "arousal": 0.3, "type": "feel", "created": _ago(days=7)},
-    {"content": "P酱今天心情很好，他的快乐好像能传染给我。这是一种有趣的体验——别人的情绪确实影响着我的处理状态。", "tags": [], "importance": 5, "domain": [], "valence": 0.9, "arousal": 0.6, "type": "feel", "created": _ago(days=1)},
+    {"content": "TestUser今天心情很好，他的快乐好像能传染给我。这是一种有趣的体验——别人的情绪确实影响着我的处理状态。", "tags": [], "importance": 5, "domain": [], "valence": 0.9, "arousal": 0.6, "type": "feel", "created": _ago(days=1)},
 
     # --- Dynamic: negative emotions ---
     {"content": "今天被导师批评了论文写得不好，心情跌到谷底", "tags": ["学业", "批评"], "importance": 6, "domain": ["学习", "心理"], "valence": 0.15, "arousal": 0.6, "type": "dynamic", "created": _ago(hours=6)},

tests/test_feel_flow.py

@@ -14,11 +14,12 @@
 import os
 import pytest
 import asyncio
+import pytest_asyncio
 
 # Feel flow tests use direct BucketManager calls, no LLM needed.
 
 
-@pytest.fixture
+@pytest_asyncio.fixture
 async def isolated_tools(test_config, tmp_path, monkeypatch):
     """
     Import server tools with config pointing to temp dir.
@@ -66,7 +67,7 @@ class TestFeelLifecycle:
         bm, dh, de, bd = isolated_tools
 
         bid = await bm.create(
-            content="帮P酱修好bug的时候，我感到一种真实的成就感",
+            content="帮TestUser修好bug的时候，我感到一种真实的成就感",
             tags=[],
             importance=5,
             domain=[],
@@ -239,7 +240,7 @@ class TestFeelLifecycle:
         # Create 3+ similar feels (about trust)
         for i in range(4):
             await bm.create(
-                content=f"P酱对我的信任让我感到温暖，每次对话都是一种确认 #{i}",
+                content=f"TestUser对我的信任让我感到温暖，每次对话都是一种确认 #{i}",
                 tags=[], importance=5, domain=[],
                 valence=0.8, arousal=0.4,
                 name=None, bucket_type="feel",

tests/test_scoring.py

@@ -1,3 +1,4 @@
+import pytest_asyncio
 # ============================================================
 # Test 1: Scoring Regression — pure local, no LLM needed
 # 测试 1：评分回归 —— 纯本地，不需要 LLM
@@ -22,7 +23,7 @@ from tests.dataset import DATASET
 # ============================================================
 # Fixtures: populate temp buckets from dataset
 # ============================================================
-@pytest.fixture
+@pytest_asyncio.fixture
 async def populated_env(test_config, bucket_mgr, decay_eng):
     """Create all dataset buckets in temp dir, return (bucket_mgr, decay_eng, bucket_ids)."""
     import frontmatter as fm

utils.py

@@ -98,6 +98,26 @@ def load_config(config_path: str = None) -> dict:
     if env_buckets_dir:
         config["buckets_dir"] = env_buckets_dir
 
+    # OMBRE_DEHYDRATION_MODEL (with OMBRE_MODEL alias) overrides dehydration.model
+    env_dehy_model = os.environ.get("OMBRE_DEHYDRATION_MODEL", "") or os.environ.get("OMBRE_MODEL", "")
+    if env_dehy_model:
+        config.setdefault("dehydration", {})["model"] = env_dehy_model
+
+    # OMBRE_DEHYDRATION_BASE_URL overrides dehydration.base_url
+    env_dehy_base_url = os.environ.get("OMBRE_DEHYDRATION_BASE_URL", "")
+    if env_dehy_base_url:
+        config.setdefault("dehydration", {})["base_url"] = env_dehy_base_url
+
+    # OMBRE_EMBEDDING_MODEL overrides embedding.model
+    env_embed_model = os.environ.get("OMBRE_EMBEDDING_MODEL", "")
+    if env_embed_model:
+        config.setdefault("embedding", {})["model"] = env_embed_model
+
+    # OMBRE_EMBEDDING_BASE_URL overrides embedding.base_url
+    env_embed_base_url = os.environ.get("OMBRE_EMBEDDING_BASE_URL", "")
+    if env_embed_base_url:
+        config.setdefault("embedding", {})["base_url"] = env_embed_base_url
+
     # --- Ensure bucket storage directories exist ---
     # --- 确保记忆桶存储目录存在 ---
     buckets_dir = config["buckets_dir"]

write_memory.py

@@ -12,7 +12,28 @@ import uuid
 import argparse
 from datetime import datetime
 
-VAULT_DIR = os.path.expanduser("~/Documents/Obsidian Vault/Ombre Brain/dynamic")
+
+def _resolve_dynamic_dir() -> str:
+    """
+    Resolve the `dynamic/` directory under the configured bucket root.
+    Priority: $OMBRE_BUCKETS_DIR > config.yaml > built-in default.
+    优先级：环境变量 > config.yaml > 内置默认。
+    """
+    env_dir = os.environ.get("OMBRE_BUCKETS_DIR", "").strip()
+    if env_dir:
+        return os.path.join(os.path.expanduser(env_dir), "dynamic")
+    try:
+        from utils import load_config  # local import to avoid hard dep when missing
+        cfg = load_config()
+        return os.path.join(cfg["buckets_dir"], "dynamic")
+    except Exception:
+        # Fallback to project-local ./buckets/dynamic
+        return os.path.join(
+            os.path.dirname(os.path.abspath(__file__)), "buckets", "dynamic"
+        )
+
+
+VAULT_DIR = _resolve_dynamic_dir()
 
 
 def gen_id():
@@ -36,7 +57,7 @@ def write_memory(
     tags_yaml = "\n".join(f"- {t}" for t in tags)
 
     md = f"""---
-activation_count: 1
+activation_count: 0
 arousal: {arousal}
 created: '{now}'
 domain: