Development skills

Configures a repo for Matt Pocock's engineering skills: issue tracker, triage labels and domain doc layout.

Builds features and fixes bugs test-first with the red-green-refactor loop.

Turns the current conversation into a spec and publishes it to the project's issue tracker, without a further interview.

Breaks a plan or spec into tracer-bullet tickets with their blocking relationships, published to the issue tracker.

Moves issues and external pull requests through triage: categorise, verify and write briefs an agent can act on.

Sets up Claude Code hooks that block dangerous git commands such as push, reset --hard and clean before they run.

Migrates test files from `as` type assertions to @total-typescript/shoehorn.

Sets up Husky pre-commit hooks with lint-staged, type checking and tests in the current repo.

Behavioural guidelines that reduce common LLM coding mistakes: avoid overcomplication, make surgical changes and state assumptions.

Reference for the Claude API and Anthropic SDKs: model ids, pricing, parameters, streaming, tool use, MCP, caching and model migration.

Guides the creation of MCP servers whose tools let an LLM work with external APIs and services.

Builds multi-component claude.ai HTML artifacts with React, Tailwind CSS and shadcn/ui, for work that needs state or routing.

Tests local web applications with Playwright: verify frontend behaviour, debug the UI, capture screenshots and read browser logs.

Migrates React projects and components from Radix UI to Base UI, including shadcn projects.

QA-tests a website or web app in a browser and returns a 1–5 quality score with evidence.

Tells Claude Code when to delegate to the cavecrew subagents for locating code, small edits and diff review, keeping output compressed.

Writes a Conventional Commits message compressed to the intent of the change.

Explores a repository read-only to locate code across files and returns path and line citations.

Reviews a diff in compressed form: one line per finding with location, problem and fix.

Diagnoses an ambiguous failure before any edit, ranking hypotheses by evidence.

Builds a feature within strict scope, reusing what the repository already has and stopping at an explicit condition.

Implements schema, data, API or dependency migrations that are reversible, with rollback and preservation proof.

Restructures code while preserving behaviour, with verification before and after each structural edit.

Fixes a bug or small behaviour change at the narrowest responsible layer, with regression proof.