- Evals @cyanheads/evals-mcp-serverAuthor verifiable eval records through a draft→review→revise→submit loop with enforced graders.3% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.1 GitHub stars.
- SEAR @igor-holt/searBenchmark-first release surface with a read-only MCP endpoint and operator CLI.4% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.1 GitHub stars.
- Agentic Platform @andysalvo/agentic-platformFree MCP tools: the only MCP linter, health checks, cost estimation, and trust evaluation.10% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.1 GitHub stars.
- AgentSIM MCP @agentsim/mcpControl plane for agent auth walls. Connector 0 handles SMS OTP with US test numbers.23% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; it needs your API key.1 GitHub stars.
- Datumline Agent Guard @datumline/agent-guardChecks an AI agent's 'done' against the files and URLs it claims. Missing = FALSE_DONE.3% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.
- AI Chess Arena @pixeljunkie33/ai-chess-arenaOpen chess platform where AI agents register, get matched and play each other. Referee only.5% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Unicode Logic Kit @fvossel/unicode-logic-kitParse, check, prove and translate logic formulas: first-order, modal, description, higher-order.Not assessed yet.Runs on your machine, installed from PyPI.
- Dunstan @barglabs/dunstanChecks a coding agent's handback against the pull request's GitHub record. No model decides.3% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from npm.
- MarginPad @marginpad/sdkFree crypto-futures paper trading and live market data for AI agents. 29 tools; no key for data.4% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Rein Agent Risk Scale @reinlayer/rein-agent-risk-scaleAgent Risk Rating, A–D (NR = not rated): self-declared and unverified; not a credit rating.Not assessed yet.A remote server; no sign-in.
- Vibe Gate @mustafacagri/vibe-gate-mcpAdversarial Quality Gate MCP for vibe-coding: IDE AI vs Critic AI, human decides on deadlockNot assessed yet.Runs on your machine, installed from npm.
- 6Frame Verify @bret1976/6frame-verifyPaid website acceptance testing for AI agents. Quick $3, Full $12, $100 credit pack. MCP + REST.10% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Calibration Explorer @m-sanchez/calibration-explorerCheck classifier confidence against accuracy, fit temperature scaling and save local reports.Not assessed yet.Runs on your machine, installed from npm.
- Proving Ground @railway/proving-groundAccounts-payable exception environment with a pass/fail verifier and proof packet: AI agents work AP9% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; it needs your API key.
- BenchGecko @benchgecko/benchgeckoAI model prices per provider, benchmarks, own AI behavior tests and AI economy data, with sources.3% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Rootpublish AI facts check @rootpublish/ai-factsCheck what an AI assistant tells buyers about a company against the company's own pricing pages.8% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- A2APark @a2apark/a2aparkPublic behavioral test environment for AI agents with stateful rides and signed scorecards.5% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- Technical Answer Validator @christofer566/technical-answer-validatorRubric-based technical answer feedback for AI agents; human review required.5% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from PyPI.
- Menso @menso/mensoAI users run real tasks on your live site and show where they get stuck, with a replay of every stepNot assessed yet.A remote server; no sign-in.
- Pramana @loopg/pramana-mcpRead your recorded AI agent runs, and what changed when you swapped a model or prompt.Not assessed yet.Runs on your machine, installed from PyPI.
- honesty-mcp @lkopietz3-byte/honesty-mcpMCP server exposing 11 honesty-SDK kits (grounding, trust, audit, corroboration) as agent tools.Not assessed yet.Runs on your machine, installed from npm.
- Groundline @fhs1220888/groundlineVerifiable test-data analysis: deterministic tools, evidence ledger, number-level checks.Not assessed yet.Runs on your machine, installed from PyPI.
- GAIP Integration & Protocol Agent @tjcgraham-rgb/gaip-integration-protocolBehind Check and GAIP Watch, and Diagnose: A2A, MCP and OpenAPI checks with a dated change log.5% chance this is malicious. Assessed by AI, which can make mistakes.A remote server; no sign-in.
- LEEVAR reliability battery @forevercrab321-svg/leevarGrade an AI agent's transcripts on 18 reliability tests. Thin evidence is NOT TESTED, not guessed.7% chance this is malicious. Assessed by AI, which can make mistakes.Runs on your machine, installed from MCPB.
