ehuerta6/cappy-code

ai-config-builder

Design or evolve reusable AI configuration such as skills, agents, rules, templates, and project instructions without creating unnecessary duplication.

grill-me

Stress-test a feature, plan, architecture decision, or product idea until the important decisions are explicit and shared.

impeccable

Design, critique, audit, and refine production frontend UI and UX with strong hierarchy, accessibility, responsive behavior, clear states, and deliberate visual decisions.

project-audit

Audit a project's AI configuration against current project needs and ai-setup to detect duplication, drift, stale instructions, and missing reusable configuration.

retro

Review completed AI-assisted engineering work to identify concrete improvements to skills, rules, context, tooling, and automated checks.

unslop

Audit and clean technical writing so it becomes concrete, concise, credible, and free of formulaic AI prose without changing supported facts.

to-issues

Break an approved spec or large feature into focused, connected GitHub Issues that can be implemented independently by fresh agents.

debug-with-evidence

Diagnose bugs and regressions by reproducing the exact failure, gathering evidence, testing hypotheses, and verifying the fix.

firebase

Work safely with Firebase projects including Firestore, Authentication, Security Rules, Emulator Suite, Cloud Functions, Storage, indexes, seeds, and trusted server access.

github-flow

Execute Git and GitHub operations using the repository's documented conventions for commits, issues, pull requests, and merging.

implement-issue

Implement one approved GitHub Issue from scope discovery through a verified, review-ready change.

issue-batch-orchestrator

Analyze multiple GitHub Issues and produce safe parallel and sequential execution waves with review gates.

review-pr

Review a pull request or completed diff separately against its requirements and against the repository's engineering standards.

verify-change

Verify a code change using the repository's actual configured checks and report concrete evidence without hiding failures.

writing-for-agents

Write and maintain skills, AGENTS.md files, agent instructions, and other documents consumed by AI agents with minimal context load and clear behavior.