MCP servers and skills: search, filter by where they run, and compare.
MCP server and CLI to evaluate RAG retrieval, agent tool calls and prompt regressions, with CI gates. No paid keys.
Node.jsMIT★ 1Few starsNo key
The plug-and-play evals layer for LLM-based projects: library, learning center, check center, API, MCP server.
SkillMIT★ 1Few starsNo key
Run LLM eval suites, cache every call, and compare two runs item by item against a noise floor. CLI and MCP server.
PythonMITFew starsNo key
Evaluation lab for adopting Model Context Protocol (MCP) for internal tooling — research briefing, open questions, and decision tracking.
MCPLink onlyFew stars
Tracking repository for our Model Context Protocol (MCP) adoption evaluation — research background, parent tracking issue, and per-area sub-issues.
MCPLink onlyFew stars
Model Finder: find the cheapest AI model that passes your own tests. MCP server + Agent Skills for model audits, evals, cheapest-first routing and local model sizing.
PythonMIT★ 4Few starsNo key
Eval and release-gate workbench for Claude agents: error analysis, code checks, an LLM judge measured against human labels.
SkillMIT★ 1Few starsNo key
Snapshot diff for agent and tool outputs: changed cases, score regressions, pass/fail verdict.
Node.jsMITFew starsNo key
Zero-dependency AST safe mathematical and boolean expression evaluator and validator preventing arbitrary code execution
MCPLink only★ 7
Standalone harness for structured-reasoning evals: register task suites, run any model callable, report accuracy, calibration, and failure taxonomy.
SkillMIT★ 1Few starsNo key
A zero-dependency, read-only CLI that checks whether your machine actually meets a project's declared requirements, and tells you exactly what to fix. Built for humans, CI, and AI coding agents (Claud
PythonMIT★ 3Few starsNo key
Paired reasoning evaluation: bootstrap uncertainty and exact discordance tests. Python research prototype with synthetic fixtures.
SkillMIT★ 1Few starsNo key
OmniEvolve: Evolutionary code & prompt optimization framework with island-model tournaments, evaluator SDK, and SQLite lineage ledger (MCP + CLI)
PythonApache-2.0★ 1Few starsNo key
Grounded reasoning, policy-checked tools, guardrails, and behavioural evals for industrial AI agents. Zero runtime deps, 119 tests, 22/22 eval cases passing.
PythonMIT★ 1Few starsNo key
Production-Grade Context Engineering Engine, MCP Server & Evidence-First Evaluation Harness for Frontier LLMs
PythonMITFew starsNo key
Research tracker for a lunch-and-learn on the Model Context Protocol (MCP): sources, findings, and workstreams.
MCPLink onlyFew stars
Research notes and tracking for our internal talk on the Model Context Protocol (MCP): origins, architecture, SDKs, and vendor adoption.
MCPLink onlyFew stars
Security briefing notes for our team's adoption of the Model Context Protocol (MCP), including supply-chain advisory highlights.
MCPLink onlyFew stars
Sourced research digest for an internal tech talk on the Model Context Protocol (MCP): overview, architecture, transports, adoption, and the latest spec revision.
MCPLink onlyFew stars
A digest of Model Context Protocol (MCP) research for the lab onboarding talk - what MCP is, its core concepts, and the current spec revision.
MCPLink onlyFew stars
Five Claude agents build fact-checked K-College lessons through an MCP server: capability tokens, tool pinning, guardrail hooks, evals (27/27 live, 67/67 MCP red-team, 216/216 parity) and token + MCP
SkillMITFew starsNo key
Multi-agent Text-to-SQL analytics agent with evals-in-CI, model routing & budgets, a sandboxed SQL executor, resilience, and observability.
SkillMITFew starsNo key
Tool-using LangGraph agent that checks a train design spec against the EU PRM TSI: cited verdicts, MCP server, FastAPI, Docker. Evaluated on tool use (verdicts, arguments, citations, escalation) with
PythonMIT★ 1Few starsNo key
Who stands behind it
