chieaid24
- 10 skills
- 0 followers
- 17 hours ago last updated
- ▌ Spec · chieaid24 bundleFull pipeline entry point for a new feature or change: interrogate the idea (sharpening domain language, writing ADRs as decisions crystallize), then route by scope - publish a [PRD] issue with child work slices for large features, or go straight to one or a few dependency-linked issues for small changes. Both converge on the same issue-creation flow. Use when starting from a rough idea toward a dependency-linked issue queue.
- ▌ Catch Up · chieaid24Daily reviewer for the dependency-aware GitHub Issues queue: reconstructs what shipped since the last run, reports active, blocked, and human-gated (HITL) work, diagnoses stalled lanes, and starts the repo's dev server for review. Read-only except for a fast-forward of a clean `main`. Use on a daily cron or on demand: catch me up, what happened yesterday, review progress, daily catch-up, or /catch-up.
- ▌ UI Audit · chieaid24 bundleAudit and repair the visual consistency of a UI, unattended. Screenshot every flow with the repo's browser runner, find defects (misalignment, overlap, clipped text, overflow) and drift from DESIGN.md or from the app's own dominant patterns, then dispatch a fix subagent per finding and re-run its probe until every flow is clean. Use when the user wants to audit the UI, fix UI inconsistencies, or screenshot every flow and check it. Not for building new screens, redesigns, or UX/copy changes - only visual-consistency repair of existing flows.
- ▌ Code Audit · chieaid24Autonomous architecture-improvement pass. Explore the codebase for deepening opportunities (shallow modules to deep modules), delegate behavior-preserving refactors to subagents, and continuously verify that nothing changed from the consumer side by keeping unit, integration, and E2E suites green. Use for unattended architecture cleanup, "audit the code", "improve the architecture", agent-driven refactoring runs, or /code-audit. Not for adding features, fixing bugs, or any change that alters observable behavior.
- ▌ Writing Style · chieaid24 bundleHouse writing style for READMEs. Use only when writing or editing a README. Not for any other prose - blog posts, release notes, announcements, docs pages, landing or marketing copy, changelogs, PR bodies, commit messages, code comments, or agent instructions (SKILL.md / AGENTS.md) - those follow the repo convention or writing-great-skills.
- ▌ Security Audit · chieaid24 bundleSecurity audit of a codebase - web apps, APIs, services, CLI tools, libraries, daemons, and more. Use when asked to find security bugs, do a security review, audit for vulnerabilities, or pen-test the code. Focuses on exploitable issues with real impact, not theoretical concerns or industry-standard behavior.
- ▌ Bootstrap Issues · chieaid24 bundleBootstrap a repo for parallel autonomous agents coordinated via a dependency-aware GitHub Issues queue - creates work-queue labels (lifecycle plus afk/hitl autonomy), a CI test gate, branch protection with self-merge, an agent issue template, a committed DESIGN.md design system for frontend repos, and the AGENTS.md/CLAUDE.md workflow docs. Use when the user wants to set up the parallel-agent / GitHub Issues flow in a repo, "bootstrap issues", port the agent workflow to a new repo, or invokes /bootstrap-issues.
- ▌ Setup Pre Commit · chieaid24Set up stack-aware pre-commit hooks - format, lint, and test before every commit. JS/TS via Husky + lint-staged + Prettier; Python via the pre-commit framework (ruff); Rust/Go via a tracked git hook (fmt + lint + test). Use when the user wants pre-commit hooks, commit-time formatting/typechecking/testing, or when bootstrap-issues wires up a repo.
- ▌ Start Next Issue · chieaid24Iteration-capped orchestrator for the dependency-aware GitHub queue -- the main agent grabs the next ready `afk` issue (skipping `hitl` issues, which need a human), drives it to a merged PR, then dispatches a fresh-context worker agent per remaining iteration (up to 3 issues total), each reporting back to the orchestrator before the next starts; any failure propagates upward and stops the run. Runs inside a git repo against that repo's queue, or from a parent folder (e.g. ~/projects) against several repos at once, listed in a `.agent-repos` file. Optionally accepts an issue number (/start-next-issue 42) or plain-text description (/start-next-issue "fix auth bug") to target a specific issue on the first iteration. Use when the user wants an agent to work a bounded batch of issues from the queue, "work the next few issues", run a capped orchestrated batch, or invokes /start-next-issue.
- ▌ Writing Great Skills · chieaid24 bundleReference for writing and editing skills well - the vocabulary and principles that make a skill predictable. Use when authoring or revising a SKILL.md, choosing model- vs user-invocation, structuring progressive disclosure, or pruning no-ops. Not for published prose a reader outside the work sees - that is writing-style.