MCP server with specialized judges to evaluate AI-generated code for security, cost, scalability, cloud readiness, and best practices.
Instructions file GitHub Copilot
Instructions for KevinRabun/judges, covering adding a new judge — step-by-step, 0. create/update agent file (agents/ .judge.md), 1. create the evaluator, 2. create the judge definition (with self-registration) and 3. add the side-effect import.
Instructions file GitHub Copilot
Instructions for KevinRabun/judges: When reviewing pull requests in this repository.
Instructions file GitHub Copilot
Instructions for KevinRabun/judges, covering release & deployment — agent instructions, files that need version updates, changelogs that need updating, 1. root changelog.md and fixed.
Instructions file GitHub Copilot
Instructions for KevinRabun/judges, covering secure coding — preventing codeql / code scanning findings, 1. regex safety (js/polynomial-redos), 2. no sensitive data in logs (js/clear-text-logging), 3. complete sanitization (js/incomplete-sanitization) and 5. cryptographic key sizes (js/insufficient-key-size).
Instructions file GitHub Copilot
Instructions for KevinRabun/judges, covering vs code extension debugging — agent instructions, step 1: get the runtime error (always do this first), step 2: classify the error, step 3: validate the manifest against official docs and step 4: fix, rebuild, and verify locally.