AI governance evidence, automated.
Mima helps engineering and compliance teams generate audit-ready evidence for EU AI Act, SOC 2, ISO 42001, and NIST AI RMF — directly from their AI systems, without manual documentation.
| Package | Description |
|---|---|
| mima-governance-python | Python SDK — runtime attestation, evidence push, policy acknowledgment |
| governance | TypeScript / Node.js SDK |
| governance-mcp | MCP server — any AI agent, any stack, 4 lines of config |
| eslint-plugin-mima | ESLint plugin — flags unattested AI calls at development time |
AI systems produce evidence records as they run — every inference call, every human-oversight decision, every prompt version change. Mima captures these records, maps them to the controls that matter (Art. 9, Art. 13, SOC 2 CC6…), and maintains a signed, auditable ledger.
When an auditor asks for evidence, you already have it.
Docs: docs.mima.works · PyPI: mima-governance · npm: @mima-ai/governance