Explore
MCP servers and skills: search, filter by where they run, and compare.
Nothing matches every word exactly, so these are the closest matches.
MCP servers and skills: search, filter by where they run, and compare.
Nothing matches every word exactly, so these are the closest matches.
Automated evaluation workflow for Agent Skills with capability extraction, test-case generation, rubric scoring, badcase analysis, DeepSeek judging, and CI reports.
SkillMITFew starsNo key
Eval harness for agent skills (SKILL.md): does the skill trigger when it should and does the model follow it - OpenAI, Qwen, DeepSeek, GLM, Kimi, any OpenAI-compatible API.
SkillMITFew starsNo key
Test whether coding agents load and follow your skills, in isolated Docker Sandboxes, traced with MLflow.
SkillMIT★ 2Few starsNo key
Transparent evaluation harness for AI agent constitutions, deterministic skill execution, and verifiable causal provenance.
SkillMIT★ 4Few starsNo key
Governed, read-only MCP access to DuckDB, Postgres and BigQuery for AI agents: semantic layer, SQL guard, row policies, PII pseudonymization, audit log, cost caps. Red-teamed and eval-gated.
PythonMITFew starsNo key
50 validated API authorization-testing skills + an expectation-matrix CLI
SkillMITFew starsNo key
CLI for Walmart Connect and Sam's Club advertising support cases — read, answer, and file them without a browser
SkillMITFew starsNo key
Local MCP server for MIDAS Gen NX / Civil NX, built on midas-nx (not an official MIDAS IT product)
PythonMITFew starsNo key
Install and manage AI agent skills from any git repo (npx skills compatible) for Claude Code, Codex, OpenCode, Gemini and more
SkillMIT★ 41No key
Evidence-backed Apache NiFi investigation for AI assistants: an independent, read-only MCP server (public preview)
PythonApache-2.0Few starsNo key
Gold drawdowns, US real yields, CFTC positioning & official reserves · 黄金宏观量化研究 · Agent Skill + Python CLI
SkillMIT★ 1Few starsNo key
Lint, security scan, test, forge and export agent skills (SKILL.md) for Claude Code, Cursor, Codex, Hermes and more. Free models supported.
SkillMIT★ 48No key
Test your Agent Skills before your agents do — lint, score, and routing-test SKILL.md files.
SkillMITFew starsNo key
An experiment: test portfolios and strategies against the S&P 500 on real data, then manage a paper or real Alpaca account with hard risk limits. Agent skills for Claude Code, Codex and Grok.
SkillMIT★ 1Few starsNo key
Universal AI Agent Skill for multi-dimensional codebase health audit, structured IMPROVEMENTS.md roadmapping, and atomic test-verified refactoring loops across modern AI agents (Claude Code, OpenAI Co
SkillMIT★ 2Few starsNo key
A reproducible observatory for discovering, validating and searching public Agent Skills.
SkillMITFew starsNo key
Derive a writing style from own text samples, adjust it with sliders and measured targets, and export it as an Agent Skill; texts in six European languages.
SkillMIT★ 1Few starsNo key
Fact-check a research paper against its own evidence: statistics, numbers, references, figures, code and reporting standards. Agent skill, CLI, MCP.
PythonMITFew starsNo key
Guided Microsoft Foundry Agent Insights onboarding with Agent Skills
SkillMITFew starsNo key
MCP server that lets AI agents search, retrieve, store, view, measure and de-identify DICOM images over both DIMSE and DICOMweb, using one set of tools
PythonApache-2.0★ 4Few starsNo key
OmniSocial | 自媒体运营全流程系统：多平台账号、素材与发布的一体化本地工作台，覆盖抖音、小红书、视频号、B 站、YouTube、TikTok 等 17 个发布目标与 Web / CLI / Skill 三层入口；uploader/ 是唯一发布引擎，默认只出计划、确认后才发，失败与未知结果如实上报。A local-first multi-platform publishing syste
SkillMIT★ 2Few starsNo key
thinkPad is a harness that turns Claude into a Socratic tutor.
SkillApache-2.0★ 2Few starsNo key
Hands-on tutorial: fast feedback loops for AI capabilities, from a coding agent skill to an Airflow Dag with no model in it
SkillApache-2.0★ 1Few starsNo key
Who stands behind it
