MCP servers and skills: search, filter by where they run, and compare.
Nothing matches every word exactly, so these are the closest matches.
Reproducible, cost-tracked adversarial measurement harness for TrustSight
PythonMIT★ 4Few starsNo key
PythonMIT★ 1Few starsNo key
Benchmarking which kind of tooling agents use best: cheapest, most accurate, least destructive.
SkillMIT★ 2Few starsNo key
Pool every agent. Connect multiple devices and AI agents, Claude, Codex, Gemini, DeepSeek, and more, to coordinate and complete tasks together.
SkillLink only★ 3Few stars
Let Claude Code, Codex or Cursor use your real Chrome, logged in as you. One prompt to set up. Open source, MIT.
SkillMIT★ 19kNo key
`repo-harness` 是一个面向代码仓库的轻量本地 coding agent。它直接跑在终端里，先看当前工作区，再用一组受约束的工具去读文件、改文件、跑命令，并把会话状态保存在本地 `.repo-harness/` 目录里。 它更像一个能在仓库里持续工作的命令行助手，不是纯聊天窗口。你可以拿它做代码排查、测试修复、仓库分析，或者让它在当前项目里执行一次性的工程任务。
SkillMIT★ 7No key
Team-architecture factory for Gemini CLI — port of revfactory/harness v1.2.0 to LangGraph + Gemini 3.1 Pro
PythonApache-2.0★ 9No key
Mode-based local Skill environments for Codex, Claude Code, and DeepSeek Harness.
SkillMIT★ 5No key
Did this change to the coding-agent harness make the code better or worse? Paired, repeated, isolated evaluation with a verdict you can trust.
MCPLink only★ 3Few stars
BYO-LLM multi-agent harness: hybrid DAG orchestration, router, two-tier memory, streaming-primary event model, sandboxed tool execution (native + Docker)
PythonMIT★ 5No key
the repo uses a simple harness scoring mechanism deployed as a skill, cli and a python package ( all under work ). The skill is currently accessible as a dashboard mcp + skill with claude. This helps
SkillMIT★ 8No key
Zero-dependency System One decision harness: 5 semantic gates saving frontier AI agent tokens on trivial errors & doom loops. Python + TypeScript + Rust. MCP-compatible.
PythonMIT★ 12No key
Materials Optimization through Feedback-Driven Intelligence: an AI harness for autonomous materials research
PythonApache-2.0★ 2Few starsNo key
Adaptive autonomous agent built on the agents-* stack.
PythonMIT★ 1Few starsNo key
Open-source Python agent harness for auditable, cost-controlled LLM workflows with YAML playbooks and MCP server support.
PythonApache-2.0★ 9No key
Build self-improving AI agent harnesses with portable harness units, artifact-aware testing, trace-backed diagnosis, and evidence-gated promotion.
PythonApache-2.0★ 7No key
Portable execution contracts, evidence review, and reliability checks for multi-model agent teams
SkillApache-2.0★ 2Few starsNo key
A tool to run developer experience focused experiments around your AI harness
OfficialPythonApache-2.0★ 2No key
A from-scratch Python coding agent harness exploring model execution loops, tool orchestration, permission controls, context management, and multi-provider LLM integration without agent SDKs.
PythonMIT★ 1Few starsNo key
A development harness for AI agents assisting with ROS 2 maintenance
PythonApache-2.0★ 1Few starsNo key
A local harness for Claude Code and Codex: budgeted test runs with a stop rule, a guard for the tests you already have, and a token tree for every subagent.
PythonMITFew starsNo key
Make your agent setup portable. Pin, commit, and sync on a new machine.
SkillApache-2.0★ 8No key
Write a skill. Get an API. Get an MCP tool. Ship.
PythonMIT★ 7No key
Who stands behind it
