Platform
Cursor
Rules, skills, and instructions for Cursor. Copy them directly or export in the platform's native format.
75 published items · Filter in browse
Saving Workspace Context
Persist project knowledge that should outlast this conversation: decisions, gotchas, research, and commands. Writes to AGENTS.md or a docs folder the project already uses. Use when the user asks to...
Responsive Testing
Open the app in the harness's browser tool at multiple viewport sizes, screenshot each, and report layout breakage. Use after UI layout changes or when the user asks to verify responsive behavior.
Release notes from merged PRs since last tag
Collects merged PRs and commits since the last tag, classifies by user impact, rewrites in user vocabulary, and adds upgrade steps.
Prompt Engineering
Write effective prompts and system prompts for current LLMs, including reasoning models, tool use, structured output, and prompt caching. Use when the user asks to improve a prompt, design a system...
PR title and body from branch diff
Generates a PR title and body with purpose-grouped changes, evidence-based testing notes, and a risks section from the branch diff.
Parallel Exploring
Explore a large codebase in parallel by launching several read-only subagents that each investigate a different area at once. Use when onboarding onto a new project, understanding architecture, or ...
Blameless incident postmortem
Blameless postmortem template with timeline anchors, TTD/TTM metrics, why-chains, owned action items, and a responder interview guide.
Safe dependency upgrade procedure
Changelog review, one-package-per-commit bumps, frozen-lockfile install, test and typecheck gates, and a written rollback plan.
Dark Mode Testing
Toggle between light and dark mode in the harness's browser tool, screenshot both states, and flag missing token mappings or contrast issues. Use after theme or styling changes, or when the user as...
Conventional Commits from a diff
Writes a Conventional Commits message from the staged diff, matching the repo's existing type and scope style.
Codebase onboarding map
Time-boxed procedure to map an unfamiliar repo (entry points, layers, entities, commands, hotspots) and persist it to a notes file.
API design review with scoring rubric
REST and GraphQL contract review checklist (naming, errors, pagination, versioning, auth) with an 8-dimension, 40-point rubric in references/.
Next.js 15 + React 19 + Vercel AI + Tailwind Cursor rule
.mdc rule for Next.js 15 App Router, React 19, Vercel AI SDK, shadcn/ui and Tailwind with a mandatory analyze-then-plan step and async request API guidance.
Next.js App Router + TanStack Query v5 Cursor rule
Scoped .mdc rule for Next.js App Router with TanStack Query v5: HydrationBoundary, Server Actions as mutations, optimistic updates, infinite scroll.
Go net/http ServeMux REST API Cursor rule
.mdc rule for building REST APIs in Go 1.22+ with the standard-library ServeMux: plan first, then idiomatic error handling, validation, logging, tests.
Definition of done: verify before reporting
Always-on rule: run tests, typecheck, lint, and build before reporting completion, and report exactly what ran.
Cursor Rule: Strict TypeScript
Auto-attached .cursor/rules/typescript.mdc that enforces strict TypeScript conventions on every .ts/.tsx file in context.
Cursor Rule: Testing Conventions
Auto-attached .cursor/rules/testing.mdc that loads when test files are in context and fixes naming, structure, mocking, and flakiness rules.
Always-on agent guardrails
Always-on rule: scope discipline, git and dependency approval, planning format, and when to ask before acting.
ConnectionGuide.txt format rule
Auto-attached rule that keeps a per-repo ConnectionGuide.txt of ports, routes, connections, and integration env vars current.
Cursor Rule: Commit and PR Hygiene
Always-applied .cursor/rules/git-hygiene.mdc covering commit scope, message format, branch names, PR descriptions, and what never to run.
Cursor Rule: API Security
Agent-requested .cursor/rules/api-security.mdc the agent pulls in when work touches auth, sessions, route handlers, or anything that reads untrusted input.
Clean Code Cursor rule
Language-agnostic .mdc rule covering naming, constants, comments, single responsibility, DRY, encapsulation, refactoring and testing hygiene.
Vercel AI SDK AGENTS.md
The AI SDK's 324-line AGENTS.md: repo map, pnpm commands, Zod and JSON-parsing rules, error-class pattern, provider architecture, changesets, per-change-type completion checklists and a Do-Not list.