Grill Me

A router that helps determine which skill or workflow to use for a given task. Use when you need guidance picking the right tool from a suite of AI agent skills.

Reviews code changes since a fixed point along two independent axes: coding standards compliance and spec/issue fidelity. Runs both reviews in parallel sub-agents and reports findings side by side. Ideal for reviewing branches, PRs, or work-in-progress changes.

A shared vocabulary and principles for designing deep, testable modules with small interfaces and high leverage. Use when designing or improving module interfaces, deciding where to place seams, or making code more maintainable and AI-navigable.

Guides systematic debugging of hard bugs and performance regressions through building a tight feedback loop, reproducing, minimizing, hypothesizing, and instrumenting. Use when encountering difficult or flaky bugs.

Actively builds and sharpens a project's domain model, challenging vague terms, stress-testing concepts with edge cases, and capturing glossaries and architectural decisions (ADRs) as they crystallize.

Craft verifiable goal prompts for AI coding agents like Claude Code, Codex, or Pi. Turns vague tasks into precise goals with checkable completion criteria, ensuring agents run unattended.

Helps plan, design, and implement A/B tests and growth experiments. Use when you need to set up split tests, create hypotheses, determine sample sizes, and analyze results for statistically valid decisions.

An accessibility expert skill for conducting WCAG compliance audits, identifying barriers, and providing remediation guidance for inclusive digital experiences.

Find and fix WCAG 2.2 accessibility issues. Operates in report mode for prioritized audits without edits, or fix mode with an audit→edit→verify loop. Uses live-DOM auditing when possible.