aoi-triple-memory-lite
par edmonddantesj
Search & ResearchRésumé
AOI Triple Memory (Lite) — file search + decision notes templates (no plugins).
Description
Ce que fait ce skill
A plugin-free memory system for AI agents using three layers: a MEMORY.md log, dated daily notes, and structured decision notes in a context/ folder. Search is handled by ripgrep across the workspace. No external databases or sync services involved.
Because it requires no embeddings database or external service, it works in any environment where ripgrep is available and leaves no sensitive data outside the local workspace.
Quand l'utiliser
- Searching past project decisions by keyword
- Logging daily work context without a plugin
- Creating tagged decision notes for architecture choices
- Reviewing what was decided about a feature last month
- Keeping a lightweight agent memory in a public or shared repo
Exemple de workflow
User asks: find notes about the royalty rail decision
- 1
Run aoi-memory search with query 'royalty rail'
- 2
ripgrep scans workspace files including MEMORY.md and context/ notes
- 3
Return matching lines with file paths and line numbers
- 4
User opens the relevant context/ note for full details
List of matching file locations and excerpts referencing the royalty rail decision
Installation
clawhub install edmonddantesj/aoi-triple-memory-liteExécute la commande CLI, ou colle l'instruction dans Claude Code, Cursor, Codex ou l'agent de code que tu utilises.