What the Responsible AI Initiative is and what it provides to the Foundation and its projects.
A proposed platform architecture to support RAI:
the llm gateway, agents, mcp, and the advisor governance layer.
The security audit pipeline: its roadmap, eval framework, and security-spec coverage (ASVS, CWE Top 25, API Top 10, ASF Baseline, SLSA).
Setup and runbooks for the systems the platform depends on: Gofannon (the agent runtime) and the vast.ai + OpenWebUI runbook (self-hosted model serving PoC).
Contributor agreement and guidelines.