All authors
treq-dev avatar

Claude Skills by treq-dev

github.com/treq-dev
9 skillsA× 8C× 10 installs0 views
App QaC

Visually verify treq UI/behavior changes by driving the real app (real jj repo via NAPI, real Rust dispatch, jsdom-rendered React) with @testing-library/user-event and capturing before/after screenshots through the Chromium rasterization harness in scripts/screenshot/. Use explicitly when the user runs /app-qa or asks to screenshot, QA, or visually check a behavior. ALSO use proactively, without being asked, right after implementing or modifying anything that changes rendered UI or user-facin...

ai-agentsjavascriptrust
0
4
Audit InterlinkingA

Audit and fix internal interlinking across the treq content site (web/learn and web/docs) for SEO and navigation. Checks existing links for relevance and broken targets, finds articles that are under-linked or orphaned, and adds links from an article's key terms to other relevant existing articles. Use when asked to audit, improve, or fix interlinking, internal links, or SEO link structure across the docs, or when asked to link a set of articles together.

ai-agentspythongo
0
4
Docs WritingA

Write or revise Treq technical documentation under web/docs/, plus product pages that document shipped app behavior (roadmap, security, settings copy). Activate whenever an agent drafts, updates, or reviews technical docs for Treq: concept pages, how-tos, tutorials, reference, security, roadmap feature status, or related sidebar/index links. Owns accuracy against the current code and product state (shipped vs WIP), prerequisite honesty, cross-page consistency, and the doc revision checklist. ...

ai-agentspythongo
0
4
Explain To MeA

Explain concepts clearly in the treq voice: direct, declarative prose for a peer engineer, grounded in Orwell's rules for plain English and ASD-STE100 (Simplified Technical English), with house banlists for buzzwords, marketing language, and AI writing tells (em dashes, "not just X but Y", rule-of-three padding, "it's important to note"). Use when the user asks to explain or clarify something, when writing or revising inline code comments, when drafting explanatory documentation, or whenever ...

ai-agentspythonrust
0
4
Pr AnnotateA

Leave inline GitHub review comments on a PR's changed files explaining why each change was made. Use automatically right after creating a pull request with mcp__github__create_pull_request (or the gh CLI), before telling the user the PR is up — do not wait to be asked. Also use when the user explicitly asks to annotate a PR's code changes or explain the "why" behind a diff.

ai-agentsgit
0
4
Service QaA

Verify treq's integration with the local Supabase CLI stack (Auth, PostgREST RPCs/RLS, Edge Functions) by driving real HTTP/RPC calls against http://127.0.0.1:54321 — no mocked supabase-js. Use explicitly when the user runs /service-qa or asks to QA auth, desktop token exchange, merge-queue RPCs, webhooks, or other Supabase-backed flows. ALSO use proactively, without being asked, right after implementing or modifying anything that changes the service contract: supabase/migrations/**, supabase...

ai-agentsgobash
0
4
Ui StoryA

Create or update Storybook stories that mount a treq React component as a presentational component, driven purely by props/args (no live jj repo, no Tauri IPC, no NAPI). Use when the user runs /ui-story, asks to "add a story", "storybook this component", "preview this component in storybook", or asks for a component to get a Cloudflare Pages preview link on its PR. Scope is app components today (src/components/**, src/components/ui/**) — the long-term goal is a components package shared betwe...

ai-agentsrustgo
0
4
WritingA

Write or revise Learn articles under web/learn/, READMEs, changelogs, and release notes. Owns Learn-site article structure, interlinking, and the doc revision pass for those surfaces. For technical product docs under web/docs/** (and roadmap/security feature-status copy), use the docs-writing skill instead. Voice, Orwell's rules, ASD-STE100, buzzword banlists, and AI-tell checks live in the explain-to-me skill: read that first whenever you draft or edit prose.

ai-agentspythongo
0
4
Agent SkillA

Work inside a Treq agent terminal: use the treq CLI to create and inspect workspaces, send files or text to the user in the Treq UI, and create commits with treq commit. Use this skill for every Treq agent session, whenever you need another workspace, should show the user a file or preview, need to record a commit, or are finishing inplace conflict resolution.

ai-agentsbashgit
0
4