# Cognitive Model Protocol (CMP)

CMP is an experimental interoperability protocol for evidence-based working models of human-AI collaboration.

Current status: `0.1.0-draft`, seeking criticism.

## Canonical Reference

Read [reference/current.md](https://github.com/filinils/cognitive-model-protocol/blob/main/reference/current.md) first. It is the canonical description of CMP.

## Repository Index

- [Agent entrypoint](https://github.com/filinils/cognitive-model-protocol/blob/main/AGENTS.md)
- [Machine-readable manifest](https://github.com/filinils/cognitive-model-protocol/blob/main/cmp.manifest.yaml)
- [Rationale](https://github.com/filinils/cognitive-model-protocol/blob/main/RATIONALE.md)
- [Draft field specification](https://github.com/filinils/cognitive-model-protocol/blob/main/SPEC.md)
- [JSON Schema](https://github.com/filinils/cognitive-model-protocol/blob/main/schema/cmp.schema.json)
- [Examples](examples.md)
- [Quickstart](quickstart.md)
- [LLM discovery](llms.txt)

## Critique CMP

The most useful contribution right now is criticism: [How to Critique CMP](critique.md).