Cursor rule Cursor
Workflow memory operating loop — read root state before acting, write state back after meaningful change.
not rated 325
changed 1mo ago A 35 tokens
cut
02Agent Claude Code
Over-engineering hunter. Extra files, wrappers, speculative types, unjustified deps. Use for /cut or "too much code".
not rated 0
changed 1mo ago A 29 tokens
original MIT
hunter
03Agent Claude Code
Logic-bug and vulnerability hunter for the current diff. Use for /hunter, a security pass, or review before a PR. Not for docs or formatting.
not rated 0
changed 1mo ago A 34 tokens
original MIT
prove
04Agent Claude Code
Skeptical verifier. Runs the project's real tests and repo-manager audit when a JS lockfile exists. Use for /prove or when the work is claimed done.
not rated 0
changed 1mo ago B 36 tokens
original MIT
astro
05Cursor rule Cursor
Astro roof. HTML first, islands explicit, version-matched docs.
not rated 0
changed 1mo ago A 298 tokens
original MIT
next
06Cursor rule Cursor
Next.js roof. Version-matched docs for THIS app. App vs Pages from the tree.
not rated 0
changed 1mo ago A 334 tokens
original MIT
pnpm
07Cursor rule Cursor
For new JS, use pnpm for add/update/run/audit. Respect an existing manager; migration needs approval.
not rated 0
changed 1mo ago B 230 tokens
original MIT
postgres
08Cursor rule Cursor
Postgres roof: parameterized SQL, forward-only migrations, measured indexes. Index only.
not rated 0
changed 1mo ago A 261 tokens
original MIT
testing
09Cursor rule Cursor
Verification invariant: proportional proof, fail closed on red, no self-grade.
not rated 0
changed 1mo ago A 410 tokens
original MIT
vite
10Cursor rule Cursor
Vite roof. SPA/SSR from this repo's plugins and config. Not Next, not Astro.
not rated 0
changed 1mo ago A 357 tokens
original MIT
debugging
11Skill Claude CodeCodex
Evidence-first debug for unknown or intermittent bugs. Use when the cause is unknown or the user asks to diagnose. Establishes evidence and distinguishes confirmed causes from hypotheses. Diagnosis alone does not authorize implementation changes.
not rated 0
changed 1mo ago A 43 tokens
original MIT
design-stack
12Skill Claude CodeCodex
Router for design turns. Picks one UI skill for the job. Use at Product Designer, @Design, UX start, or any visual UI pass.
not rated 0
changed 1mo ago A 33 tokens
original MIT
Skill Claude CodeCodex
Marketing, landing, and campaign pages (Elaya). Product apps use premium-ui-craft instead.
not rated 0
changed 1mo ago A 24 tokens
original MIT
Skill Claude CodeCodex
Linear/Stripe/Apple-grade UI craft. Use for Product Designer / @Design, dashboards, shadcn, or generic SaaS UI.
not rated 0
changed 1mo ago A 34 tokens
original MIT
Skill Claude CodeCodex
Audit then fix generic or AI-looking sites in place. Product apps apply findings through premium-ui-craft.
not rated 0
changed 1mo ago A 27 tokens
original MIT
testing
16Skill Claude CodeCodex
TDD order, mocks, gauntlet, regression naming. Use when writing or expanding tests, not because testing.mdc is on. Produces the verifying test for this change; does not authorize unrelated refactors.
not rated 0
changed 1mo ago A 46 tokens
original MIT
supabase
17Cursor rule Cursor
Supabase roof: tenant isolation, service role, pooler. Index only.
not rated 0
changed 1mo ago A 243 tokens
original MIT
core
18Cursor rule Cursor
Engineering roof: loop, craft, architecture, ladder, types, complexity, stack, harness.
not rated 0
changed 1mo ago A 866 tokens
original MIT
handoff
19Skill Claude CodeCodex
Schema-validated session continuity. Use when work spans chats, the user asks to pause or hand off, or the next session must resume without the transcript. Writes state/handoff.json. Not authority for new goals.
not rated 0 1mo ago A 46 tokens
original MIT
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: