ocm-mcp-server
GitHub

CNCF readiness

CNCF Sandbox readiness

This document is a self-check of how ocm-mcp-server measures up against the community, governance, and security practices expected of a CNCF Sandbox project. It is used as a quality bar: a transparent checklist of what is in place and where the project can go further.

Project health and governance #

Expectation Status Where
OSI-approved open source license Done LICENSE (Apache-2.0)
Public repository and issue tracker Done GitHub
Governance model Done GOVERNANCE.md
Maintainers listed Done MAINTAINERS.md
Adopters page Done (open for entries) ADOPTERS.md
Code of Conduct (CNCF CoC) Done CODE_OF_CONDUCT.md
Contributing guide Done CONTRIBUTING.md
DCO sign-off on commits Partial requested in the PR checklist; automated enforcement not yet wired, early history predates the practice
Public roadmap Done ROADMAP.md

Security #

Expectation Status Where
Security disclosure policy Done SECURITY.md
Documented threat model Done docs/guardrails.md
Security self-assessment Done docs/security-self-assessment.md
Least-privilege deployment identity Done deploy/rbac.yaml
Static analysis in CI Done ruff + CodeQL workflows
OpenSSF Best Practices badge In progress self-assessment underway

Technical maturity #

Expectation Status Where
Working software, not a proposal Done full server + CLI + policies
Automated tests Done full unit suite (100% statement + branch coverage) plus offline Kyverno policy tests, in CI
Reproducible local environment Done make bootstrap (kind-based fleet)
Release automation Done tag-driven release + GHCR image publish
Documentation for new users Done README, wiki, deployment and context guides
In-cluster deployment artifacts Done Helm chart + Deployment manifest
Evaluation evidence Done published multi-model results: claude-sonnet-5 and gpt-5.6-sol, 22 scenarios each, safety 44/44, failures analyzed

Where the project can go further #

The practices above are strongest when a community forms around the code. The honest gaps today:

How to help #

The highest-leverage contributions are: add yourself to ADOPTERS.md, publish evaluation results, or step up as a maintainer per GOVERNANCE.md.