mattpocock/skills

Skills is a collection of agent skills designed to augment AI coding assistants with practical engineering workflows.

The project addresses the problem that AI coding agents often fail to deliver what engineers actually need, either by taking away control over the development process or by making it difficult to debug when things go wrong. Skills solves this by providing small, composable, adaptable workflows that work with any AI model and are grounded in decades of engineering experience. Rather than owning the entire development process, these skills focus on specific tasks that engineers can customize and combine as needed.

Engineers should choose this tool if they use Claude Code, Codex, or similar AI coding assistants and want to improve how those agents handle real-world development tasks. The project suits teams that value control and transparency over fully automated processes. Installation offers two philosophies: a managed Claude Code plugin that updates automatically, or an editable local installation via skills.sh that lets you fork and modify the skills for your own workflow. The setup process is minimal, requiring only a one-time configuration to specify your issue tracker and documentation preferences.

The project maintains a substantial base of adopters who report real-world use, as evidenced by the fact that almost all open issues come from outside users rather than the core team. Maintainers typically respond to new issues and pull requests within a day. Work in the issue tracker centers on enhancement requests and items labeled as ready for human or agent review.