We build polyglot, .NET-first software with an AI-assisted development workflow. Our public repositories are currently the engineering platform behind that work: shared quality standards, CI building blocks, isolated runners, and the agent tooling we develop with.
- standards — shared repository-quality configs and engineering/review conventions, distributed to consuming repos for consistent human and AI contributions.
- ci-workflows — SHA-pinned, configurable composite actions (one per code-quality tool) and reusable GitHub Actions workflows, aggregated by consumers into a single status gate.
- ci-runner — an official Actions runner image plus a native Windows scale-set controller. Each job gets a fresh, one-job Docker worker, with GitHub-hosted runners as fallback.
- Marketplaces of reusable, repo-agnostic agent tooling — skills, agents, hooks, and MCP servers — one per assistant: claude-code-plugins, codex-plugins, and cursor-plugins.
- Shared standards make the expected quality bar explicit for people and agents.
- Reusable workflows enforce that bar consistently across repositories.
- GitHub governance is managed as infrastructure-as-code so settings are reviewable and reproducible.
- AI assists implementation and review; automated checks and maintainer judgment remain the release gates.
- Browse the repositories above, and open an issue or pull request where you see something to improve.
- Review our Code of Conduct and contribution guidelines before contributing.
- Found a security issue? Follow our security policy and do not report vulnerabilities publicly.