Memtrace Continuous Memory

Overview

Keep the knowledge graph live while editing. watch_directory triggers incremental re-indexing on file saves (~80 ms typical latency after debounce).

Required parameters — watch_directory

{ "path": "/abs/path/to/repo", "repo_id": "memdb" }

Full parameter spec for every Memtrace tool: references/mcp-parameters.md (bundled at the memtrace-skills plugin root).

Steps

1. Confirm indexed

list_indexed_repositories()

Run index_directory first if missing (poll check_job_status; stop after ~5 minutes and report the job id — see memtrace-index).

2. Start watching

See JSON above. Returns immediately; watcher runs in background.

3. Confirm active watches

list_watched_paths()

Response shape: see Output.

4. Edit normally

Next find_symbol / get_symbol_context call sees saved changes after debounce (500 ms) + incremental persist (80 ms).

5. Stop watching (kill switch)

unwatch_directory:

{ "path": "/abs/path/to/repo" }

Idempotent — unwatching an already-unwatched path is a no-op.

Latency expectations

Output

watch_directory returns an immediate acknowledgment; the watcher runs in the background. list_watched_paths returns:

{
  "watches": [
    { "path": "...", "repo_id": "...", "branch": "...", "started_at": "...", "origin": "..." }
  ],
  "count": 1
}

There is no persist_ms field in this response.

Common Mistakes