MCP Tools DirectoryAbout this integration
Self-curating agent memory kept honest by survival-based selection over measured outcomes.
- Transport
- stdio
- Authentication
- none
- Initial setup
- none
- Runtime
- unattended
- Evidence
- documented
- Version
- 0.7.1
- Package
- darwin-memo
- Last compatibility test
- Not independently tested
Connect your agent
darwin-memoPublisher documentation reviewed; package and MCP server were not executed. The MCP surface can mutate the selected memory store through add, settle, forget, pin, unpin, abandon, and tick operations. Optional cloud-LLM integrations and their keys are separate from the documented local MCP launch.
Capabilities: Query a persisted agent lesson store with provenance, Record and settle measured outcomes against memory entries, Inspect, add, forget, pin, unpin, and advance stored memory state
Connected profiles
Additional details
io.github.rogermsc/darwin-memo
Source ↗ · Checked 2026-09-170.7.1
Source ↗ · Checked 2026-09-17CC0-1.0; package licenses are separate
Source ↗ · Checked 2026-09-17pypi
Source ↗ · Checked 2026-09-17darwin-memo
Source ↗ · Checked 2026-09-19unspecified
Source ↗ · Checked 2026-09-19The local MCP process operates on the configured memory file, and its full state and open tickets persist across process sessions and restarts.
Source ↗ · Checked 2026-09-19The documented local stdio MCP launch uses a memory-file argument and no service credential; the core runs offline. Optional Anthropic and compatible cloud clients are separate extras, not prerequisites for this connection.
Source ↗ · Checked 2026-09-19Install the mcp extra and configure the existing stdio connection to launch darwin-memo-mcp with an explicit --memory path.
Source ↗ · Checked 2026-09-19Publisher documentation reviewed; package and integration endpoint not executed or independently security-audited.
Source ↗ · Checked 2026-09-19