{"name":"io.github.wjabanjj/aifp-mcp","slug":"wjabanjj-aifp-mcp","title":null,"description":"Persistent memory MCP server for AI coding assistants. Chinese-first and fully local.","url":"https://mcp.market/server/wjabanjj-aifp-mcp","rating":null,"grade":"C","score":60,"certified":false,"status":"active","category":"other","tags":[],"presence":{"score":28,"stars":3,"forks":0,"downloads_week":239,"last_push_at":"2026-08-19T11:50:28.000Z","license":"NOASSERTION"},"uptime":null,"claimed":false,"transport":"npm","callable_via_gateway":false,"default_price_micros":0,"repository":"https://github.com/wjabanjj/aifp-mcp","website":null,"version":"1.5.8","remotes":[],"packages":[{"registryType":"npm","identifier":"aifp-mcp","version":"1.5.8","transport":{"type":"stdio"},"environmentVariables":[{"description":"Optional LLM API key for automatic memory recognition (COGNITION_LLM_API_KEY or ANTHROPIC_API_KEY)","format":"string","isSecret":true,"name":"COGNITION_LLM_API_KEY"}]}],"tools":[{"name":"batch_delete","description":"批量删除记忆（软删除，可恢复）。需要清理不再需要的多条记忆时调用，比逐条删除高效。","write_action":true,"price_micros":0,"input_schema":null},{"name":"batch_update","description":"批量更新记忆的标签或重要性。需要统一调整多条记忆的分类或权重时调用。","write_action":true,"price_micros":0,"input_schema":null},{"name":"consolidate_memories","description":"手动触发记忆巩固：根据使用频率和重要性自动晋升记忆层级（episodic → internalized → growth）。系统启动时也会自动运行，通常无需手动调用；需要立即巩固时可用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"deduplicate_memories","description":"按内容前 80 字分组去重，保留 salience 最高的。当记忆库出现重复时调用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"derive_memories","description":"从对话中识别并保存长期记忆。本地运行（需配置 COGNITION_LLM_API_KEY 或 ANTHROPIC_API_KEY）。","write_action":false,"price_micros":0,"input_schema":null},{"name":"diffuse_memories","description":"记忆扩散搜索——从一组记忆出发，沿关联关系向外多跳扩散，发现间接相关的知识。需要连接服务器进行加权扩散（按 relation_type 区分权重）；本地模式不可用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"explain_query","description":"解释检索结果路径分解：展示每条记忆来自哪些路径（FTS/向量/实体/时间/关键词）以及分数构成。当需要理解\"为什么这条记忆被检索到\"、排查检索质量时调用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"export_memories","description":"导出记忆为 JSON，支持按类型/层级筛选。需要备份、迁移或分析记忆数据时调用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"export_memories_md","description":"把记忆导出为 Markdown 笔记（带 frontmatter），写入指定目录——配合 Obsidian 等笔记软件使用，在 Obsidian 里能看到全部记忆。当用户想把记忆同步到 Obsidian vault 时调用。本地可用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"find_perception_path","description":"查找两个记忆节点之间的最短因果路径（双向 BFS）。需要连接服务器进行深度搜索；本地模式不可用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"flush_recognizer","description":"手动触发 Recognizer 立即处理待识别队列。通常由后台调度器自动运行，但需要立即落库时可用此工具。","write_action":false,"price_micros":0,"input_schema":null},{"name":"get_confidence_stats","description":"查看置信度分布统计：展示当前记忆库中各置信度层级（0.3/0.5/0.7/0.9）的数量分布。当需要评估记忆库整体可信度时调用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"get_memory","description":"按 ID 获取单条记忆的完整详情（内容、类型、标签、置信度等）。当已通过 search_memories 或 recall_context 拿到记忆 ID，需要查看该条记忆的完整信息时调用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"get_memory_tree","description":"获取记忆的层级树结构（按 parent_id 组织的父子关系）。当需要理解记忆之间的从属/包含关系时调用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"get_perception_graph_stats","description":"获取因果图统计信息——节点数、边数、关系类型分布、中心节点。需要连接服务器；本地模式不可用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"get_related_memories","description":"获取与指定记忆关联的其他记忆（Hebbian 共现关联），即\"一起出现\"的记忆。当需要围绕一条记忆展开关联信息、发现间接主题时调用。本地模式可用（基于本地共现矩阵）；连接服务器后计算更完整。","write_action":false,"price_micros":0,"input_schema":null},{"name":"get_stats","description":"获取记忆感知系统统计信息（记忆总数、类型分布、层级分布等）。当需要了解系统整体状态、记忆库规模时调用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"get_top_experiences","description":"获取按衰减分数排序的 Top N 经验教训。当需要提取记忆库中最重要/最常被引用的经验时调用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"get_user_profile","description":"获取用户画像——聚合记忆中关于用户本人的偏好、事实、习惯类信息。当需要了解用户的完整背景（喜欢什么、不喜欢什么、身份事实、习惯）时调用，比 search_memories 更全面；AI 建议/结论类记忆不会混入画像。本地可用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"list_memories","description":"分页列出记忆库中的记忆，按重要性倒序。当需要浏览记忆库全貌、盘点已记住的内容、或按类型/层级筛选查看时调用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"merge_memories","description":"合并重复记忆：将多条内容重复/相似的记忆合并到一条目标记忆。当检索结果中发现明显重复的记忆时调用，减少冗余。","write_action":false,"price_micros":0,"input_schema":null},{"name":"observe_turn","description":"记录对话内容到记忆识别队列。当本轮对话中用户透露了可能值得记住的偏好、事实、决策时，可调用此工具让系统自动判断是否值得入记忆。轻量快速（<5ms），仅入队不阻塞。","write_action":false,"price_micros":0,"input_schema":null},{"name":"recall_context","description":"一键回忆（比 search_memories 更全面）：当需要全面了解某个主题的来龙去脉时调用此工具。本地模式：仅 FTS5 全文 + 向量语义基础检索（不含感知链/关联/扩散）。连接服务器后：额外叠加感知链深度溯源（backward depth=8）+ Hebbian 关联 + 多跳扩散。返回结果含 source 字段。","write_action":false,"price_micros":0,"input_schema":null},{"name":"reimport_sources","description":"导入记忆来源：扫描指定目录（如 Obsidian vault 路径）或 COGNITION_SOURCES 配置的目录，把 .md/.json 等文件导入记忆库（frontmatter 自动剥离）。内容哈希去重不重复。用户新增了笔记后调用此工具同步。","write_action":false,"price_micros":0,"input_schema":null},{"name":"rotate_observation_logs","description":"清理 7 天前的观察日志文件。定期维护用，一般由系统自动处理，无需手动调用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"save_memory","description":"保存一条记忆。当用户明确告知关于他自己的新信息（偏好、事实、经历、观点、项目决策等），且你觉得值得长期记住时，调用此工具。内容会自动建立全文索引和向量索引，后续搜索可命中。","write_action":false,"price_micros":0,"input_schema":null},{"name":"scan_memory_patterns","description":"扫描近期记忆模式：按类型/标签聚类分析，了解当前记忆库的内容分布和趋势。当需要总结用户近期关注点、记忆主题变化时调用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"scan_observation_patterns","description":"扫描最近 N 小时的观察日志，返回高频错误模式。当需要从操作日志中提取失败规律时调用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"search_memories","description":"搜索记忆库，查找与查询词相关的历史记忆。当用户提及之前讨论过的话题、项目、偏好、技术决策，或需要了解用户背景信息时，主动调用此工具。支持关键词 + 语义双路检索，适合大部分回忆场景。本地模式：FTS5 全文 + 向量语义 6 路检索。连接服务器后：额外启用 Z-score 融合排序 + 冲突证据补充（自动拉入矛盾记忆的另一端）。返回结果含 source 字段标识 \"server\" 或 \"local\"。","write_action":false,"price_micros":0,"input_schema":null},{"name":"session_mine","description":"挖掘对话会话中的决策、踩坑、编辑循环，提炼经验。当需要从一段对话历史中提取可复用经验时调用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"share_memory","description":"设置记忆的共享状态（允许其他 Agent 访问）。多 Agent 协作场景下，需要把某条记忆共享给其他 Agent 时调用。","write_action":false,"price_micros":0,"input_schema":null},{"name":"trace_perception_chain","description":"追踪记忆的感知链（BFS），查看\"为什么\"和\"导致什么\"。需要连接服务器（depth=8 双向 BFS）；本地模式不可用。返回结果含 depth 和 relation 字段。","write_action":false,"price_micros":0,"input_schema":null},{"name":"validate_memory","description":"验证记忆内容质量（长度/瞬态/猜测三重门禁）。保存记忆前可调用它预检内容是否值得入记忆。","write_action":false,"price_micros":0,"input_schema":null}],"scan":{"score":60,"grade":"C","scanned_at":"2026-11-03T19:42:25.072Z","report":{"scannerVersion":"0.1.10","scannedAt":"2026-11-03T19:42:24.958Z","components":{"code":{"score":15,"max":25,"notes":["36 source files scanned"]},"reliability":{"score":-1,"max":20,"notes":["no gateway calls yet and no remote to probe"]},"poisoning":{"score":-1,"max":15,"notes":["tools not inspected (local package is not executed); not counted"]},"auth":{"score":6,"max":15,"notes":["static API keys via environment variables"]},"maintenance":{"score":12,"max":15,"notes":["last push 76 days ago"]},"identity":{"score":6,"max":10,"notes":["registry namespace matches repository owner"]}},"findings":[{"id":"install.script","severity":"medium","component":"code","title":"npm install lifecycle script present"},{"id":"install.script","severity":"medium","component":"code","title":"npm install lifecycle script present","evidence":"package.json: …\"tsc\", \"prepublishOnly\": \"tsc\", \"postinstall\": \"node scripts/postinstall.mjs\", \"bu…"}],"inputs":{"packages":[{"registryType":"npm","identifier":"aifp-mcp","version":"1.5.8","found":true,"license":"SEE LICENSE IN LICENSE","hasInstallScripts":true,"dependencyCount":3,"publishedAt":"2026-08-15T07:09:34.284Z","repositoryUrl":"git+https://github.com/wjabanjj/aifp-mcp.git","weeklyDownloads":239}],"repo":{"found":true,"owner":"wjabanjj","repo":"aifp-mcp","archived":false,"pushedAt":"2026-08-19T11:50:28Z","stars":3,"forks":0,"openIssues":0,"ownerType":"User","ownerAvatarUrl":"https://avatars.githubusercontent.com/u/271795639?v=4","ownerCreatedAt":"2026-03-28T15:58:05Z","license":"NOASSERTION"},"icon":{"url":null,"source":"none"},"presence":{"stars":3,"forks":0,"downloadsWeek":239,"license":"NOASSERTION","lastPushAt":"2026-08-19T11:50:28.000Z","score":28}}}},"grade_history":[],"reviews":[]}