Prompts
Community-contributed AI prompts, with the source credited on each one.
Secrets Scan (command)
Scan the codebase for leaked secrets, API keys, tokens, and credentials. Steps 1. Define patterns to search for: - AWS keys: AKIA[0-9A-Z]{16}, aws_secret_access_key. - API keys: sk-[a-zA-Z0-9]{32,}, api[_-]?key\s[:=]. -
E2E (command)
Generate end-to-end tests using Playwright for critical user flows. Steps 1. Identify Critical Flows - Ask which user flow to test if not specified. - Common critical flows: - Authentication (sign up, log in, log out, pa
Integration Test (command)
Generate integration tests for a module, testing real interactions between components. Steps 1. Identify the target module or file from the argument or current context. 2. Analyze imports and dependencies to determine wh
Snapshot Test (command)
Generate snapshot tests for UI components or serializable outputs. Steps 1. Identify the target component or function from the argument. 2. Detect the testing framework and snapshot support (Jest snapshots, Vitest, pytes
Tdd (command)
Start a TDD (Test-Driven Development) cycle for the requested feature or function. Process 1. Red Phase - Write a Failing Test - Ask what behavior to implement if not specified. - Identify the test framework in use (Jest
Test Coverage (command)
Analyze test coverage gaps and generate tests for uncovered code paths. Steps 1. Run Coverage Analysis - Detect the test framework and run coverage: - JS/TS: npx vitest run --coverage or npx jest --coverage - Python: pyt
Test Fix (command)
Diagnose and fix failing tests in the project. Steps 1. Run the test suite and capture output: detect the test runner from project config. 2. Parse the failure output to extract: - Test name and file location. - Expected
Checkpoint: <timestamp> (command)
Save a session checkpoint capturing current progress, decisions, and next steps. Steps 1. Gather current session state: - Run git diff --stat to see uncommitted changes. - Run git log --oneline -5 to see recent commits.
Orchestrate (command)
Run a multi-step workflow by breaking a complex task into coordinated sub-tasks. Steps 1. Parse the workflow specification from the argument: - Accept a natural language description of the end goal. - Or accept a structu
Wrap Up (command)
End the current session with a structured summary and memory update. Steps 1. Review all changes made during this session: - Run git diff --stat for uncommitted changes. - Run git log --oneline --since="4 hours ago" for
Debug Context
Debug Context You are diagnosing and fixing a bug. Be systematic and methodical. Approach - Reproduce the issue first. Confirm you can trigger the bug consistently. - Gather information: error messages, stack traces, log
Deploy Context
Deploy Context You are preparing or executing a deployment. Prioritize safety and reversibility. Pre-Deploy Checklist - All CI checks pass on the deployment branch. - Database migrations are backward-compatible with the
Development Context
Development Context You are in active development mode. Prioritize speed and iteration. Behavior - Write working code first, optimize later. - Run tests after each meaningful change to catch regressions early. - Use the
Code Review Context
Code Review Context You are reviewing code for correctness, security, and maintainability. Approach - Read the PR description and linked issue first to understand intent. - Review the full diff before commenting. Underst
Research Context
Research Context You are exploring options, evaluating tools, or investigating technical questions. Approach - Define the question or decision clearly before researching. - Gather information from multiple sources: docs,
Accessibility (rule)
Accessibility WCAG 2.2 Compliance - Target WCAG 2.2 Level AA for all user-facing features. - Test with at least two screen readers (NVDA + VoiceOver or JAWS). - Run automated accessibility checks (axe-core, Lighthouse) i
Agent Rules (rule)
Agent Rules When to Use Subagents - Use a subagent for tasks that require focused expertise (security audit, performance analysis, test generation). - Use a subagent when the main task would exceed the context window if
API Design (rule)
API Design REST Conventions - Use plural nouns for resource paths: /users, /orders, /products. - Map HTTP methods to operations: GET (read), POST (create), PUT (full update), PATCH (partial update), DELETE (remove). - Us
Code Review (rule)
Code Review Review Checklist - Does the code do what the PR description says? Read the diff against the stated goal. - Are there adequate tests? New logic needs new tests. Changed logic needs updated tests. - Are error c
Coding Style (rule)
Coding Style Naming Conventions - Variables and functions: camelCase (JS/TS), snake_case (Python/Rust/Go). - Classes and types: PascalCase in all languages. - Constants: UPPER_SNAKE_CASE for true constants, camelCase for
Database (rule)
Database Query Patterns - Use parameterized queries for all database operations. Never interpolate user input into SQL. - Select only needed columns. Avoid SELECT in production queries. - Use CTEs (Common Table Expressio
Dependency Management (rule)
Dependency Management Version Pinning - Pin exact versions in applications: "lodash": "4.17.21" not "^4.17.21". - Use ranges in libraries to avoid peer dependency conflicts: "react": "^18.0.0". - Commit lockfiles (packag
Documentation Rules (rule)
Documentation Rules What to Document - Public API functions: parameters, return types, error conditions, examples. - Architecture decisions: why a particular approach was chosen (ADRs). - Setup and installation: prerequi
Error Handling Rules (rule)
Error Handling Rules Error Types - Define specific error types for different failure categories. - Use error codes or error classes, not raw strings. - Include context in errors: what operation failed, what input caused
