MCP servers and skills: search, filter by where they run, and compare.
Nothing matches every word exactly, so these are the closest matches.
Hermes-native Cursor SDK/ACP/stream-json harness plugin
SkillMIT★ 5No key
One control plane for Skills, Agents, MCP servers, commands, hooks, and permissions across AI harnesses.
PythonMIT★ 3Few starsNo key
Per-project AI coding harness for Claude Code · Codex. Profiler + 10-dim interview build a different harness for every project.
SkillMIT★ 4Few starsNo key
Iterative agent harness improvement: run a coding agent on a hard task, generate the reusable tooling it was missing, qualify it, and replay fresh sessions with it activated. Works with Codex and Clau
PythonApache-2.0★ 48No key
RLE: harness × model benchmark on a live RimWorld colony. Swappable agent harnesses (Felix multi-agent, coding agents over MCP, or your own plugin) are scored identically against an unmanaged baseline
PythonMIT★ 9No key
harness-run is a Python library for running coding agents locally and in remote sandbox environments through a common API
PythonApache-2.0★ 6No key
Coordination layer for CC, Codex, Gemini CLI and other harnesses
SkillApache-2.0★ 8No key
AIBOM generator for AI agent harnesses (Hermes, OpenClaw) - CycloneDX 1.6 output for runtime, model, skill, MCP, hook, and secrets-surface components
PythonMIT★ 3Few starsNo key
mini-harness 是一个分阶段演进的 LLM Agent 框架（Harness），目标是从内核出发，逐步补全一个商业级 Agent 系统所需的全部基础设施
PythonMIT★ 3Few starsNo key
Turn agent development into deterministic state, evidence, and gate-controlled delivery for humans and AI.
SkillApache-2.0★ 3Few starsNo key
Answer a few questions → download a production-ready agent harness for Claude Code, Codex, or Cursor. Bilingual (EN/KO), Docker-ready.
PythonMIT★ 5No key
Dynamic, on-the-fly generated harnesses for AI agents. The LLM designs its own runtime per task.
SkillMITFew starsNo key
Package manager for AI agent assets — 42 skills, 5 CLIs, 1 plugin. Skills for AI/LLM agent engineering, eval-driven dev, spec-driven dev, database (MySQL/PG/Redis/Kafka/Mongo/Oracle/NL2SQL), K8s/Docke
PythonApache-2.0★ 25No key
Universal harness layer for AI coding agents — one command (uvx harnessforge init) sets up your repo for Claude Code, Cursor, Codex, Gemini CLI, Aider, and any other coding agent.
SkillMIT★ 3Few starsNo key
GUI-HARVEST: Self-Improving GUI Agents through Evidence-Driven Harness Evolution
SkillApache-2.0★ 5No key
🔌 Turn harness internal tools into a standard MCP server — compatible with Claude Code, Codex, OpenClaw, and more.
MCPLink only★ 1Few stars
🌐 官网：harnessjob.com Harness Job 是一个基于 Claude Code Skill 架构的智能求职助手，通过 CDP 浏览器自动化 + AI 评分 + 人工确认的三层策略，帮助开发者高效求职。
SkillLink only★ 3Few stars
A verifiable multi-agent harness for mathematical modeling with deterministic gates, execution receipts, artifact DAGs, evidence grounding, fresh review, and human-in-the-loop controls.
SkillMIT★ 4Few starsNo key
Local-first personal AI context loop for Code Agent conversations
SkillMIT★ 3Few starsNo key
Attractor Before Harness 是一个面向 AI 协作开发的收敛框架与 CLI 工具集。项目核心思想来自"先定义系统要收敛到哪里，再用 harness 持续纠偏"的方法论：先把吸引子、基线、计划、验证、审计和记忆显式化，再让开发过程围绕这些对象运行。
SkillLink only★ 4Few stars
AI workflow optimization tool for Claude Code - session continuity, feature tracking, context management, and subagent delegation
PythonMITFew starsNo key
Forge your own agent harness — a config-to-code generator that renders a thin, standalone Python agent harness you fully own. No LangChain/LangGraph/ADK lock-in, no dependency on the generator after g
PythonMIT★ 21No key
Build durable agent harnesses so smaller models can do repeatable, verifiable work — a versioned folder with tools, guardrails, validators, and measured evidence from every run.
SkillApache-2.0★ 4Few starsNo key
MCP server for Harness - Continuous delivery and cloud cost management
MCPLink only★ 2Few stars
Who stands behind it
