AgentSkillsCN

Memory

内存

SKILL.md

Memory

Semantic search over your notes (MEMORY.md, memory/*.md) and chat history (every conversation is automatically indexed). "Remember what we said yesterday?" pulls from logs—no manual step.

Chat history baked in — Not just files. Conversations are stored in workspace/chat-log/YYYY-MM-DD.jsonl and embedded so memory_search finds them. Use memory_search for queries like "what did we decide about X?", "what did I ask yesterday?", "remember when we talked about Y?"

Auto-indexing — Every message you send gets embedded and added to the index. No manual "moo index" or sync needed. File-based notes (MEMORY.md, memory/*.md) are synced when you run memory_search or on watch; chat is indexed as you talk.

Tools (pass tool in arguments: "memory_search" or "memory_get")

  • memory_search — Set tool: "memory_search", query (required). Optional: maxResults, minScore. Semantically search notes and chat history. Returns snippets with path and line range (paths may be MEMORY.md, memory/2025-02-15.md, or chat-log/2025-02-16.jsonl).
  • memory_get — Set tool: "memory_get", path (required, from memory_search). Optional: from, lines. Read a snippet by path (including chat-log/*.jsonl for past conversations).

Config

  • Add "memory" to skills.enabled. Set an embedding API key (e.g. OpenAI) in .env; if omitted, the first LLM model's key is used.
  • Workspace: ~/.cowcode/workspace/. Create MEMORY.md and optionally memory/*.md. Chat logs live in workspace/chat-log/ and are created automatically.