Every Read, Edit, and Bash that Claude makes turns into a compressed observation. Observations get summarized at session end. Relevant ones get auto-injected into future prompts so the next session starts with context from the last one — no re-explaining the codebase, no re-discovering decisions.

Memory injection starts on your second session in a project.

The first session in a fresh project seeds memory; subsequent sessions receive auto-injected context for relevant past work. Run if you want to front-load the entire repo into memory in a single pass (~5 minutes, optional).

Everything stays in ~/.claude-mem on this machine.

Nothing leaves your machine except calls to whichever AI provider you configured for compression (Claude / OpenRouter / Gemini). The SQLite database, vector index, logs, and settings all live under that directory and are removed cleanly on .