Can Myco work with LangChain?
Yes. Teams can integrate Myco as memory infrastructure while still using LangChain orchestration.
Comparison
Myco vs LangChain: Build-It-Yourself Memory Stack vs Ready-Made Memory Layer
LangChain is a strong framework for assembling agent systems. Myco gives teams a ready-made, deterministic memory layer, self-hosted on your own Postgres, without stitching every component manually.
| Feature | Myco | LangChain |
|---|---|---|
| Primary scope | Ready-made memory layer | Agent framework |
| Fact extraction | Deterministic write path | LLM-based |
| Provenance | First-class via brain_why | Partial |
| Shared memory | Native Postgres source of truth | Depends on app wiring |
| Cross-session recall | Built into the product | Best effort |
| Import ChatGPT/Claude history | mycobrain-ingest --from chatgpt/claude-export | — |
| Read-only REST for non-MCP consumers | mycobrain-rest (search + why, loopback) | — |
| Data portability | Plain Postgres tables | Framework shaped |
Bottom line
Choose LangChain for maximal framework flexibility. Choose Myco when you want production memory capabilities without building every retrieval and governance layer yourself.
FAQ
What does Myco remove from the build burden?
Myco removes most memory plumbing: a deterministic write path with content-hash deduplication, first-class provenance via brain_why, and graph-backed recall.