Judges Panel
A panel of 45 specialized judges that evaluate code across security, cost, compliance, accessibility, and architectural dimensions. Built for AI-generated code review, it combines deterministic AST analysis with LLM-powered deep reviews, exposing tools like evaluateCode, applyFixes, and generateReport through the MCP interface. Ships with 200+ auto-fix patches, SARIF output for GitHub Code Scanning, and agentic skills like ai-code-review and release-gate. The CLI companion lets you run judges eval from CI or terminal, while the MCP server brings the full tribunal into Claude, Cursor, or any MCP client. Includes diff-only evaluation for PRs, license scanning for supply chain compliance, and baseline suppression for known findings.
- GitHub stars
- 6
- tools
- 10
- transport
- STDIO
- last updated
- Apr 7, 2026
Tools
10 toolsPublic tool metadata for what this MCP can expose to an agent.
evaluate_v2Run a V2 context-aware tribunal evaluation with policy profile calibration, context ingestion, runtime evidence hooks, specialty feedback aggregation, and confidence scoring.
Run a V2 context-aware tribunal evaluation with policy profile calibration, context ingestion, runtime evidence hooks, specialty feedback aggregation, and confidence scoring.
No parameter schema in public metadata yet.
evaluate_app_builder_flowRun a 3-step app-builder workflow: tribunal review, plain-language translation of top risks, and prioritized remediation tasks.
Run a 3-step app-builder workflow: tribunal review, plain-language translation of top risks, and prioritized remediation tasks.
No parameter schema in public metadata yet.
evaluate_public_repo_reportClone a public repository URL, run the full judges panel across eligible source files, and generate a consolidated markdown report.
Clone a public repository URL, run the full judges panel across eligible source files, and generate a consolidated markdown report.
No parameter schema in public metadata yet.
get_judgesList all available judges with their domains and descriptions.
List all available judges with their domains and descriptions.
No parameter schema in public metadata yet.
evaluate_codeSubmit code to the full judges panel, where all 45 judges evaluate independently and return a combined verdict.
Submit code to the full judges panel, where all 45 judges evaluate independently and return a combined verdict.
No parameter schema in public metadata yet.
evaluate_code_single_judgeSubmit code to a specific judge for targeted review.
Submit code to a specific judge for targeted review.
No parameter schema in public metadata yet.
evaluate_projectSubmit multiple files for project-level analysis with all 45 judges plus cross-file architectural analysis.
Submit multiple files for project-level analysis with all 45 judges plus cross-file architectural analysis.
No parameter schema in public metadata yet.
evaluate_diffEvaluate only the changed lines in a code diff by running all 45 judges on the full file and filtering findings to specified lines.
Evaluate only the changed lines in a code diff by running all 45 judges on the full file and filtering findings to specified lines.
No parameter schema in public metadata yet.
analyze_dependenciesAnalyze a dependency manifest file for supply-chain risks, version pinning issues, typosquatting indicators, and dependency hygiene.
Analyze a dependency manifest file for supply-chain risks, version pinning issues, typosquatting indicators, and dependency hygiene.
No parameter schema in public metadata yet.
evaluate_git_diffEvaluate only changed lines from a git diff, using either a live repository path or pre-computed unified diff text.
Evaluate only changed lines from a git diff, using either a live repository path or pre-computed unified diff text.
No parameter schema in public metadata yet.
Sponsors
Agent CaffeineSponsoredClose your lid, come back to finished work
SnugSponsoredThe clipboard manager for the AI eraPromote your product hereReach 125k+ visitors monthly
WaxSponsoredBlock anything with supervision, without erasing your iPhonePromote your product hereReach 125k+ visitors monthly
knobsSponsoredSee your site on any phone, any day, in any languageReadme
Judges Panel
An MCP (Model Context Protocol) server that provides a panel of 45 specialized judges to evaluate AI-generated code — acting as an independent quality gate regardless of which project is being reviewed. Combines deterministic pattern matching & AST analysis (instant, offline, zero LLM calls) with LLM-powered deep-review prompts that let your AI assistant perform expert-persona analysis across all 45 domains.
Highlights:
- Includes an App Builder Workflow (3-step) demo for release decisions, plain-language risk summaries, and prioritized fixes — see Try the Demo.
- Includes V2 context-aware evaluation with policy profiles, evidence calibration, specialty feedback, confidence scoring, and uncertainty reporting.
- Includes public repository URL reporting to clone a repo, run the full tribunal, and output a consolidated markdown report.
- 200+ deterministic auto-fix patches (see
src/patches/index.ts) plus LLM-powered deep review.
🧪 Many commands in
printHelpare experimental/roadmap. By default, we show GA commands only. SetJUDGES_SHOW_EXPERIMENTAL=1to reveal stubs; these may not be wired yet.
Details
- Repository
- github.com/kevinrabun/judges
- Website
- kevinrabun.github.io/judges
- Category
- Security & Pentesting
- Package
@kevinrabun/judges- Transport
- STDIO
- Registry
- MCP registry, active
More from kevinrabun
Related Security & Pentesting MCP servers
View all
Skill Audit Mcpeltociear/skill-audit-mcpMCP server: static security scanner for MCP servers, agent skills & plugins. 17 attack patterns.5stars
npm Registry MCP Serveryawlabs/npmjs-mcpnpm registry MCP server — package intelligence, security audits, dependency analysis5stars
Argusironclawdevs27/argusAutomated web QA via Chrome DevTools MCP — a11y, security, visual, env diff, CI; Aegis redaction.5stars
Service Public Franceonenicolas/mcp-service-publicFrench public services: tax, property, admin, education, healthcare, security, risks, legal texts5stars
Excalidrawdebu-sinha/excalidraw-mcp-serverSecurity-hardened Excalidraw MCP server with auth, rate limiting, and 14 tools3stars