Darwin MCP Memory

This repository stores the stateful memory layer for Darwin MCP.

Purpose

• Keep long-lived skill metadata and species definitions outside the stateless brain service.
• Provide a single source of truth for what species exist and how they are registered.

Structure

• dna/registry.json: canonical registry of registered species.
• species/: Python species implementations loaded by the brain.
• requirements.txt: Python dependencies needed by species.

Typical Workflow

1. Update species code in species/.
2. Update dna/registry.json if species registration metadata changes.
3. Commit and push changes in this memory repository.
4. In the parent repository, commit the updated submodule pointer.

Notes

• Treat dna/registry.json as authoritative state.
• Keep species files focused and testable.