$ show mattpocock/skills
This repository is a meta-application: a curated, opinionated framework of prompts ('skills') designed to govern the behavior of AI coding agents like Claude Code and GitHub Copilot Codex. It does not contain traditional application code but instead defines a structured software development methodology executed by an AI agent. The system transforms an AI agent from a simple chat interface into a disciplined, stateful development partner that plans, implements, reviews, and maintains code according to explicit workflows. The architecture is a monolith of interdependent markdown files organized into a skill hierarchy. The `ask-matt` skill acts as a router, directing user requests to the appropriate workflow. Core primitives like `grilling` (structured interview) and `domain-modeling` (glossary/ADR creation) are reused by higher-level skills. Execution engines like `implement` and `code-review` drive the actual development cycle. A key innovation is the 'stateful workspace' pattern, where the agent writes decisions to files (GLOSSARY.md, ADRs, specs) to maintain context across sessions and manage token limits. The codebase demonstrates exceptionally high craftsmanship in prompt engineering, with clear patterns for context management, sub-agent delegation, and human-in-the-loop interaction. It is actively maintained with a detailed changelog. However, its complexity is a double-edged sword: the tight coupling between skills and the reliance on specific model behavior creates a fragile ecosystem. The documented open bugs and high cognitive load for users are significant risks, preventing it from achieving an 'exceptional' rating despite its innovative design. The repository is a brilliant but brittle system for professional AI-assisted development.
The LLM identified 6 files worth highlighting for reuse:
The core interview primitive. It defines a structured loop of rounds, frontier management, and fact/decision extraction. It is explicitly designed as a model-invoked skill, making it a highly reusable component for any skill that needs to gather information from the user systematically.
A self-contained skill for building a shared domain vocabulary (GLOSSARY.md) and recording architectural decisions (ADRs). It solves the critical problem of context drift in AI agents by persisting shared understanding to the filesystem. It is invoked by planning skills like `grill-with-docs`.
A novel pattern for generating interactive bash scripts that guide a human through manual procedures. The `template.sh` provides a robust framework with stages, progress indicators, and secure input handling. It is a completely self-contained, reusable component for automating complex setup or deployment steps.
A structured debugging methodology that forces the agent to build a 'tight feedback loop' (a failing test or command) before attempting a fix. The `hitl-loop.template.sh` is a reusable bash script that implements this loop. It is a non-obvious, highly effective pattern that prevents aimless debugging.
Implements a 'Two-Axis Review' pattern, running Standards and Spec reviews in parallel sub-agents. This provides a structured, repeatable quality gate that is decoupled from the implementation skill. It is a self-contained component that can be applied to any codebase.
The meta-skill for authoring other skills within this framework. It defines the core patterns (context pointers, leading words, information hierarchy) that make the entire system work. It is a self-contained guide to effective prompt engineering for AI agents.