leek-skills

Run a bounded architectural refactor loop toward a stated design.

Drive an existing wayfinder map, spec, or set of tickets to completion through fresh top-level Claude Code, Codex, or Grok sessions.

Test a web app from a start URL and a plain-language prompt. A planner splits the prompt into cases, cheap parallel browser workers run them, and you get one verdict table plus any page errors. Read-only unless the prompt asks for a write and you confirm it.

Sweep the web for new chatter on a topic, skip what earlier runs already saw, verify what is left, and write one ranked, cited findings file to feed a blog post or a digest email.

Three-axis review of the diff since a fixed point: Correctness (bugs, failure modes, security), Standards (repo conventions) and Spec (ticket/PRD fidelity). Use when the user wants to review a branch, PR, WIP changes, or asks to "review since X".

Audit a codebase, path, or branch diff for classic code smells against the 56-smell Luzkan catalog (PHP/Laravel and TS/React heuristics). Use when the user asks for a code-smell audit or smell-check of a module or diff.

Shared vocabulary for designing deep modules in a PHP/Laravel codebase. Use when the user wants to design or improve a class or module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill (to-spec, implement, architecture-satisfaction) needs the deep-module vocabulary.

Commit the work done in this session, skipping files a parallel session also touched.

Run a dependency audit for Composer or npm projects and propose a safe upgrade plan.

Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.

Mine recent AI-coding session logs for reusable patterns and propose concrete improvements.

Build and sharpen a project's domain model: challenge terms against the glossary, resolve fuzzy language, stress-test concepts with concrete scenarios, and record CONTEXT.md entries and ADRs the moment decisions land. Use when pinning down domain terminology or a ubiquitous language, recording an architectural decision, or when another skill (grilling, wayfinder, to-spec) needs the domain model maintained.

Read a branch's latest failed CI runs, triage each failure from its existing logs, and fix the real ones.

Verify, triage, fix, and close review-bot GitHub issues one finding at a time.

Triage open Laravel Nightwatch production issues by root cause, fix the legit ones, and resolve them with evidence.

Triage active PostHog error-tracking issues by root cause, fix the legit ones, and resolve or suppress the rest with evidence.

Run a grilling session: the user wants their plan, decision, or idea stress-tested one question at a time.

Run a grilling session with domain-modeling alongside, capturing terms in CONTEXT.md and decisions as ADRs while they land. The default front door for planning work in an existing codebase.

Interview the user one decision at a time until reaching shared understanding, resolving each branch of the decision tree with recommended options and trade-offs. Use when a plan, design, or idea needs its soft spots found and forced into the open, or when another skill needs the interview protocol.

Compact the current conversation into a handoff document for another agent to pick up.

Run a conservative code-project housekeeping pass: one proven low-risk cleanup at a time.

Implement one Laravel work item end to end: scope and claim it when trackable, TDD at chosen seams, commit, review, verify, and resolve it.

Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.

Create, list, or remove git worktrees for Laravel projects served by Laravel Herd on macOS. Use when the user needs a branch worktree, .test URL, or per-worktree PHP/composer isolation.

Convene a panel of agent CLIs (grok, claude, codex, agy) on a prompt or a code review, then report one consensus grouped by how much they agree.

Build a throwaway prototype to answer a design question. Use when the user wants to sanity-check whether a state model or logic feels right, or explore what a UI should look like.

Commit this session's work, bump the version, tag, push, and publish a GitHub release with grouped notes.

Audit and clean Git repository state: branches, PRs, stashes, worktrees, and the local databases old worktrees left behind.

Investigate a question against high-trust primary sources and capture the findings as a cited Markdown file. Use when the user wants a topic researched, docs or package facts gathered, reading legwork delegated, or a wayfinder research ticket resolved.

Triage, fix or rebut, then reply-and-resolve every review comment on a pull request, including AI-review bot findings (Codex, Copilot, Gemini). Use when the user wants to address, respond to, action, or clear the review comments on a PR.

Resolve an in-progress git merge or rebase conflict while preserving both sides' intent. Use when a merge/rebase is blocked on conflicts, or the user asks to finish a conflicted merge or rebase.

Sweep a dusty .scratch/: file research into docs/, remove finished efforts, and report the work still open.

Report what is still open across .scratch/: every effort's open tickets and decisions, what blocks it, how stale it is, and the next skill to run. Read-only — never files research, removes anything, or touches disk. Companion to scratch-cleanup, which does those.

Configure this repo for the engineering skills: the markdown tracker layout, triage status vocabulary, and domain doc layout.

The red → green loop tuned for Pest/PHPUnit in a Laravel codebase: seams, what a good test is, and the anti-patterns to refuse. Use when building features test-first or when another skill needs the loop.

Teach the user a new skill or concept, within this workspace.

Turn a decision the user can't fully answer into a Markdown questionnaire for someone else to fill in, async or over a meeting.

Turn the current conversation into a spec (PRD) and save it as markdown under .scratch/: no interview, just synthesis, with Laravel test seams chosen from the tdd ranking rules rather than asked about. Use when the user wants a spec or PRD written from what was just discussed, when a grilling or wayfinder session has settled its decisions, or when an existing spec needs revising in place.

Break a spec into tracer-bullet vertical-slice tickets with explicit blocking edges, take one green light on the breakdown, and save one markdown file per ticket beside the spec.

Move markdown issues and external-PR requests through triage roles: categorise, verify, grill if needed, and write agent-ready briefs.

Triage a GitHub pull request end to end, from reviews and checks through fixes to a clean merge.

Exercise a change end to end in the running application (hit the route, run the command, click through the page) and report what actually happened. Use after tests pass, when a change needs proving in the real app, or when another skill's build step ends.

Stop. That last message did not land. Re-pitch it.

Plan work too big for one agent session as a map of decision tickets in markdown under .scratch/, then resolve them one per session until the route to the destination is clear.

Not sure which skill to reach for? Name the situation and get routed to one.

Generate an interactive bash wizard that walks a human through steps only they can perform. Use when provisioning infrastructure, creating credentials or CI secrets in a third-party dashboard, or running a one-off migration or cutover. For steps the agent can perform itself, perform them.

Reference for writing documents an agent consumes. Use when creating or editing a skill, or when modifying AGENTS.md, CLAUDE.md, or a reference doc reached by a pointer.