メニュー
Search persistent project memory across sessions — decisions, bugs, features, discoveries
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Show token / tool usage stats from the local telemetry log. Use when you want to know "which tools am I burning context on", "which skills are expensive", or "was yesterday's session mostly Read/Grep or actually productive".
Parallel quality audit with 7 specialized agents (Opus). Finds bugs, violations, and quality issues. Use audit for fixes, brainstorm for features.
Manage environment variables with Doppler — auto-install CLI, login, link projects, wrap commands with `doppler run`. Replaces scattered .env files with a hub/spoke architecture.
Scaffolds new projects or onboards existing ones. Detects stack, creates monorepo/single-app, configures strict tooling. Use for greenfield or first-time setup.
Archives completed stories from prd.json to reduce token usage.
Autonomous task execution with testing and security. Works through all tasks without stopping.
| name | mem-search |
| description | Search persistent project memory across sessions — decisions, bugs, features, discoveries |
| triggers | ["mem search","mem recent","mem decisions","mem bugs","mem timeline","mem stats","remember","what did we","what was","last session","past sessions"] |
| allowed-tools | Bash, Read |
| model | opus |
| user-invocable | true |
Search the project's persistent memory database. Observations are captured automatically during sessions.
| Say | Does |
|---|---|
mem search <query> | Keyword search across all observations |
mem recent | Last 10 observations for this project |
mem decisions | All architectural/design decisions |
mem bugs | All bug fixes |
mem timeline <query> | Session-level view with summaries |
mem sessions | List all past sessions |
mem stats | Memory database statistics |
Memory is stored in SQLite at ~/.claude/auto-dev-memory.db. Observations are captured automatically by the PostToolUse hook and classified by type:
mem search — returns titles + timestamps only (~50-100 tokens)mem timeline — shows session context around matches (~500 tokens)This 3-layer approach saves ~10x tokens vs dumping full context.
Run queries via the memory-db CLI:
# Search
node ~/.claude/scripts/memory-db.js search "$(pwd)" "auth middleware"
# Recent observations
node ~/.claude/scripts/memory-db.js recent "$(pwd)" 10
# Decisions only
node ~/.claude/scripts/memory-db.js decisions "$(pwd)"
# Bug fixes only
node ~/.claude/scripts/memory-db.js bugs "$(pwd)"
# Session history
node ~/.claude/scripts/memory-db.js sessions "$(pwd)"
# Timeline search
node ~/.claude/scripts/memory-db.js timeline "$(pwd)" "database"
# Stats
node ~/.claude/scripts/memory-db.js stats "$(pwd)"
Content wrapped in <private>...</private> tags is automatically stripped before storage. Secrets, API keys, and sensitive data in private tags never reach the database.