MCAF is an open framework built by developers who believe AI coding agents need structure, not just prompts. This page recognizes the people and projects that influenced it.
- Managed Code — Framework design and development
- Geoffrey Huntley — Creator of the Ralph Loop (Ralph Wiggum technique) and loop engineering. His work on iterative agent execution and verification informed the Ralph Loop used across MCAF's skills and agent instructions. See his original article, Ralph Wiggum as a "software engineer".
- Roger Johansson — His post about self-learning code agents inspired the self-learning logic that became the core of MCAF. His broader OSS work through Asynkron, and his habit of turning hard engineering problems into practical tools, also pushed MCAF toward sharper, more executable workflows instead of vague prompt folklore.
- Asynkron — Thank you for building practical OSS that treats architecture, diagnostics, and AI-assisted engineering as things that should be executable, inspectable, and useful in the real world.
- Asynkron.QuickDup — Roger Johansson's AI-era structural clone detector helped shape how MCAF thinks about duplication review: fast candidate generation first, then deliberate human and AI validation before refactoring.
- Asynkron.Profiler — Roger Johansson's profiling work reinforced the idea that diagnostics should be scriptable, reviewable, and friendly to terminal-first AI workflows instead of being trapped in GUI-only tooling.
- Christopher Bradford — His human-review prompt strategy for large AI-generated code drops inspired the new review-planning skill and the idea of prioritizing review by user flow and risk
- Oksana Kovalchuk — Her idea to require explicit acceptance criteria before implementation helped sharpen the MCAF workflow so agents define observable outcomes first, then write tests against those criteria. Also on GitHub.
- Agent Skills — We’re grateful for this open specification (originally developed by Anthropic) that made “skills” concrete: file-based packaging, progressive disclosure, and practical integration patterns
- AGENTS.md — Thank you for standardizing the idea of a repository-native instruction file; this is a critical foundation for how MCAF treats agent rules as versioned, reviewable code
- Microsoft code-with-engineering-playbook — Its GitHub playbook informed the engineering reference material now packaged into the MCAF baseline skills
Want to contribute? Open a PR or issue on GitHub.