Checkmate is a Spec Oracle Engine built for AI-driven development. It verifies that code changes made by autonomous agents actually deliver what your spec promised using real evidence.
# CheckMate | [](https://www.npmjs.com/package/checkmateai) [](https://github.com/richardsondx/checkmate/stargazers) By [@Richardsondx](https://x.com/richardsondx) [](https://youtu.be/fM8DmUkt7WY) AI-powered specs verification that block bad code, see through hallucination, and prevent AI from breaking your code. CheckMate is an AI Test Driven Development tool that challenges AI's overconfidence with a logical reasoning verification process using LLM. It's built with Cursor AI in mind. ## Requirements CheckMate works best with both API keys: - **OpenAI API Key**: Required for GPT models (default `quick` verifier) - **Anthropic API Key**: Required for Cursor models (default `reason` generator) ## Quick Setup ```bash # Install globally npm install -g checkmateai # Initialize in your project npx checkmateai init ``` That's it! `checkmate init` takes care of everything - creating directories, config files, and setting up Cursor integration. It will setup your app and generate a checkmate rules folder with all the necessary rules: These rules will include: - pre-task.mdc - Runs before each task - post-task.mdc - Runs after each task - post-push.mdc - Runs after each push - spec-assistant.mdc - Helps with spec creation and format guidance - spec-linter.mdc - Automated linting and fixing of spec files - verification-trigger.mdc - Triggers feature verification workflow - autofix-enforcer.mdc - Enforces auto-fix attempts on failures - drift-detector.mdc - Detects spec-vs-code drift - non-interactive-mode.mdc - For headless CheckMate runs - ai-feature-validation-guidelines.mdc - Instructional guide for AI validation - ai-verify-llm-reasoning-workflow-docs.mdc - Documentation for LLM reasoning workflow After
Agent that generates comprehensive documentation, API references, architecture diagrams, and developer onboarding guides from existing code.
Agent configuration for systematic bug investigation that traces issues from error logs through the codebase to root cause with suggested fixes.
Agent for integrating third-party APIs including SDK setup, type generation, error handling, retry logic, and rate limit management.
Cursor's built-in autonomous coding agent that can make multi-file edits, run terminal commands, search the codebase, and iteratively build features with minimal human intervention.
Cloud-based autonomous coding agent that runs in the background on remote sandboxed environments, handling complex multi-step tasks while you continue working.
Cursor's multi-file editing agent within Composer mode that can create, edit, and delete files across your entire project in a single conversation.