Cursor's config: skills, rules, subagents, commands
Overview
Comprehensive configuration for Cursor IDE to maximize development productivity. This repository provides agents (prefix agt-), skills (prefix skill-), custom commands, and development rules organized by domain and language. This configuration repository is designed to be linked/synchronized to .cursor/ directories in your projects. It provides: - : Specialized AI assistants (prefix agt-); domain dev (general) or qa (QA/Playwright); no -assistant suffix - : Reusable knowledge (folder prefix skill-) organized by language and domain - : Custom commands for development workflows - : Development guidelines and conventions - : agt-dev-* (general dev: commit, test, security, logging, frontend quality, etc.), agt-qa-* (QA/Playwright) - : All skill folders start with skill- (e.g. skill-conventional-commits, skill-qa-playwright-maintain-conventions) Link or copy this repository to your project's .cursor/ directory: Type /agt- in Cursor chat to list all agents.
README
Cursor Configuration Repository
Comprehensive configuration for Cursor IDE to maximize development productivity. This repository provides agents (prefix agt-), skills (prefix skill-), custom commands, and development rules organized by domain and language.
Overview
This configuration repository is designed to be linked/synchronized to .cursor/ directories in your projects. It provides:
- Agents: Specialized AI assistants (prefix
agt-); domaindev(general) orqa(QA/Playwright); no-assistantsuffix - Skills: Reusable knowledge (folder prefix
skill-) organized by language and domain - Commands: Custom commands for development workflows
- Rules: Development guidelines and conventions
Prefixes
- Agents:
agt-dev-*(general dev: commit, test, security, logging, frontend quality, etc.),agt-qa-*(QA/Playwright) - Skills: All skill folders start with
skill-(e.g.skill-conventional-commits,skill-qa-playwright-maintain-conventions)
Structure
.cursor-configs/
├── AGENTS.md # Index by prefix + link instruction
├── README.md # This file
│
├── agents/ # all agt-*.md here (no subdirectories)
│ # agt-dev-dotnet-logging.md, agt-dev-commit.md, agt-qa-playwright-maintain.md, etc.
│
├── skills/
│ ├── dotnet/ # skill-correlation-id-tracking, skill-opentelemetry-instrumentation
│ ├── nodejs/ # (same)
│ ├── frontend/ # skill-frontend-instrumentation-otel
│ ├── shared/ # skill-testing, skill-conventional-commits, skill-security, etc.
│ └── qa/ # skill-qa-playwright-*, skill-frontend-qa-friendly
│
├── commands/
│ └── wb-commit.md # Commit workflow command
│
├── rules/ # Development rules and guidelines
│
└── docs/
└── CREATING_ASSETS.md # Guide for creating new assets
Quick Start
1. Link to Your Project
Link or copy this repository to your project’s .cursor/ directory:
# Option 1: Symlink (recommended)
ln -s /path/to/.cursor-configs .cursor
# Option 2: Copy
cp -r /path/to/.cursor-configs/* .cursor/
2. Available Agents
Type /agt- in Cursor chat to list all agents. Filter by domain: /agt-dev- (general dev), /agt-qa- (QA/Playwright).
Invoking an agent: In Cursor chat, use / (slash) + agent name + your task. Example: /agt-qa-playwright-add-flow + “Add login flow for the UI”.
Dev — .NET
- agt-dev-dotnet-logging: Structured logging with correlation-ID tracking
- agt-dev-dotnet-opentelemetry: OpenTelemetry instrumentation
Dev — Node.js
- agt-dev-nodejs-logging: Structured logging with correlation-ID tracking (Node.js)
- agt-dev-nodejs-opentelemetry: OpenTelemetry instrumentation (Node.js)
Dev — Frontend (Vite / TypeScript)
- agt-dev-frontend-instrumentation: Frontend OpenTelemetry (Grafana/Tempo/Mimir); instrument new pages or review instrumentation code (Vite 3 + TypeScript)
- agt-dev-frontend-quality: Frontend test-friendly for QA (accessibility, stable selectors; for dev use)
Dev — Shared (multi-language)
- agt-dev-commit: Conventional Commits, context grouping, GitHub CLI
- agt-dev-test: Suggests/creates unit and integration tests
- agt-dev-security: Security vulnerabilities, OWASP Top 10
- agt-dev-refactor: Refactoring, SOLID, code quality
- agt-dev-documentation: JSDoc, XML docs, README, API docs
- agt-dev-idempotency: Idempotent operations
- agt-dev-jira-update: Safely updates Jira cards (Atlassian MCP)
QA (Playwright / frontend testability)
- agt-qa-playwright-maintain: Maintains QA Playwright project conventions
- agt-qa-playwright-add-test: Adds new API or UI test
- agt-qa-playwright-add-data: Adds/extends test-data (inputs, builder)
- agt-qa-playwright-add-fixture: Adds/extends fixture in fixtures/index.ts
- agt-qa-playwright-add-flow: Adds new test flow (folders, specs, test-data)
- agt-qa-playwright-context: Answers about structure, conventions, docs (read-only)
3. Using Agents
Agents are automatically available when this configuration is linked to .cursor/. In chat, invoke an agent with / + agent name + prompt (e.g. /agt-qa-playwright-add-flow + task description). Examples:
/agt-dev-commit Suggest commits for my changes
/agt-dev-test Create tests for this function
/agt-dev-frontend-instrumentation Instrument this new Vue page with usePageTelemetry
/agt-qa-playwright-maintain Align this spec with project conventions
4. Available Skills
Skills (folder prefix skill-) are referenced by agents. Organized by language and domain:
- dotnet/: skill-correlation-id-tracking, skill-opentelemetry-instrumentation
- nodejs/: (same)
- frontend/: skill-frontend-instrumentation-otel
- shared/: skill-testing, skill-security, skill-code-quality, skill-conventional-commits, skill-idempotency, skill-jira-update, skill-performance
- qa/: skill-qa-playwright-maintain-conventions, skill-qa-playwright-add-new-test, skill-qa-playwright-add-test-data, skill-qa-playwright-add-fixture, skill-qa-playwright-add-new-flow, skill-frontend-qa-friendly
5. Custom Commands
- wb-commit: Ensures commits follow Conventional Commits pattern with context-based grouping. Uses GitHub CLI (
gh) for PR creation.
Use in Cursor chat:
/wb-commit
The command:
- Groups changes by context/module
- Creates commits following Conventional Commits format
- Uses
git addandgit commitfor local operations - Uses
gh pr createfor PR creation (notgit push) - Protects
mainbranch (requires explicit authorization)
How It Works
Cursor Detection
Cursor automatically detects:
- Agents in
.cursor/agents/(flat; no subdirectories — Cursor only lists agents at this level) - Skills in
.cursor/skills/(including subdirectories) - Commands in
.cursor/commands/ - Rules in
.cursor/rules/
Path References
All paths in agents and skills use the full path from .cursor/ (with skill- prefix for skill folders):
.cursor/skills/dotnet/skill-correlation-id-tracking/SKILL.md.cursor/skills/shared/skill-conventional-commits/SKILL.md.cursor/agents/agt-dev-commit.md
Creating New Assets
See docs/CREATING_ASSETS.md for detailed guides on:
- Creating new agents (prefix
agt-, domaindevorqa, no-assistantsuffix) - Creating new skills (folder prefix
skill-) - Creating new commands
- Best practices and templates
Language Support
Currently supported:
- .NET (C#): Full support with agt-dev-dotnet-* agents and skill-* skills
- Node.js (TypeScript/JavaScript): Full support with agt-dev-nodejs-* agents and skill-* skills
- Frontend (Vite 3 + TypeScript): agt-dev-frontend-instrumentation, agt-dev-frontend-quality; skill-frontend-instrumentation-otel
- Multi-language: Shared agt-dev-* agents and skill-* skills
- QA / Playwright: agt-qa-* agents and skill-qa-* / skill-frontend-qa-friendly
Contributing
When adding new assets:
- Put agents directly in
.cursor/agents/(no subdirs); skills stay inskills/dotnet/,skills/nodejs/,skills/frontend/,skills/shared/, orskills/qa/ - Use prefix
agt-for agents (name without-assistant); use prefixskill-for skill folder names - Use proper frontmatter; reference skills using full paths (
.cursor/skills/.../skill-/SKILL.md) - Update this README and AGENTS.md if adding major features
References
- Cursor Documentation
- Conventional Commits
- AGENTS.md - Minimal root configuration
Recommended Tools
Try a different keyword or remove a filter.
Install
npx skillfish add whitebeardit/.cursor