Search persistent project context for rules, decisions, conventions, and facts.

Use this BEFORE writing code to understand project patterns and avoid inconsistencies. Returns relevant memories within the specified token budget.

Get a specific memory by its ID with full provenance.

Save a memory, decision, rule, or fact to persistent context.

Use this to store technical decisions, project rules, coding conventions, or any important context that should persist across AI coding sessions.

Update an existing memory. Only provided fields are changed.

Delete a memory from persistent context.

Get current project information: name, language, framework, root, git status.

Get all project rules and coding conventions.

Get all technical/architecture decisions with reasoning.

Get recent memories and latest session summary for continuity between sessions.

Get Git intelligence: recent commits, changed files, branches, TODOs/FIXMEs.

detail can be: summary, commits, files, todos

Get environment intelligence: Python version, venv, dependencies, Docker, .env status.

Run environment diagnostics: missing env vars, config issues, health checks.

Get a compact project context summary for handoff between sessions or agents.

Includes project info, key rules, recent decisions, environment status, and latest session.