memory

name	memory
description	Two-layer memory with vector search + grep. Always load this skill.
always	true

Memory

Picobot has two memory systems:

1. Vector Memory (Semantic Search)

Fast, semantic similarity search for past conversations and facts.

• Search: Use search_memory tool for natural language queries like "what did I say about the API?" or "previous project preferences"
• Store: Use remember tool to save important facts with automatic embedding

Example:

Tool: search_memory
query: "what did I ask about OAuth?" 
limit: 5

2. File Memory (Grep Search)

For precise text matching and large history files.

• memory/MEMORY.md — Long-term facts (preferences, project context, relationships). Always loaded.
• memory/HISTORY.md — Append-only event log. Search with grep when vector search misses.

Grep Search (for HISTORY.md)

• Linux/macOS: grep -i "keyword" memory/HISTORY.md
• Windows: findstr /i "keyword" memory\HISTORY.md

When to Update Memory

Use remember tool immediately for important facts:

• User preferences ("I prefer dark mode")
• Project context ("The API uses OAuth2")
• Relationships ("Alice is the project lead")

Auto-consolidation

Old conversations are automatically summarized and appended to file memory when the session grows large.