Best evals & quality skills for agents
A ranked comparison of evals & quality skills, MCP servers and repos for AI agents: 94 of 427 passed our security scan. Ranked by curation, then GitHub stars. No paid placement.
- 1. Superpowers (obra) Skill, 301k GitHub stars
- 2. Matt Pocock Skills (handoff, research, grill-me) Skill, 295k GitHub stars
- 3. Anthropic Skills (skill-creator, webapp-testing) Skill, 182k GitHub stars
ItemScanGitHub starsLicense
Superpowers (obra)Repo pass301kMIT Matt Pocock Skills (handoff, research, grill-me)Repo pass295kMIT Anthropic Skills (skill-creator, webapp-testing)Repo pass182k– Mem0Pass67kApache-2.0 Humanizer (anti-AI-slop)Repo pass58kMIT Marketing Skills (Corey Haines)Repo pass55kMIT Obsidian skillsRepo pass50kMIT CCXTPass44kMIT agent-browser (Vercel Labs)Repo pass44kApache-2.0 Playwright MCP (Microsoft)Repo pass38kApache-2.0 LangfusePass36kNOASSERTION Task Master (claude-task-master)Repo pass28kNOASSERTION Beads (issue memory for agents)Pass28kMIT promptfooPass26kMIT vectorbtPass9.4kNOASSERTION backtesting.pyPass9.1kAGPL-3.0 Firecrawl MCP Server (official)Repo pass7.6kMIT Exa MCP Server (official)Repo pass5.1kMIT Notion MCP server (official)Repo pass4.7kMIT prediction-market-analysisPass3.9kMIT AI Marketing Skills (Eric Siu / Single Grain)Repo pass3.6kMIT Google Analytics MCP (official)Repo pass3.5kApache-2.0 Google Workspace MCPRepo pass3.4kMIT Supabase MCP (official)Repo pass3.0kApache-2.0 Web quality skills (Lighthouse / Core Web Vitals)Repo pass3.0kMIT Financial Datasets MCP serverRepo pass2.3kMIT pmxt (unified prediction-market API)Pass2.2kMIT Stripe AI (official MCP + skills)Repo pass2.0kMIT Keeper.sh calendar sync + MCPRepo pass1.4kAGPL-3.0 GTM engineer AEO/GEO skillsRepo pass1.3kMIT Alpaca MCP server (official)Repo pass1.0kMIT Sentry MCP / toolkit (official)Repo pass930NOASSERTION Kraken CLI (official, built-in MCP)Pass761MIT SEO Skill (Search Console MCP + CLI)Repo pass596Apache-2.0 Oracle3 prediction-market agentRepo pass475Apache-2.0 samber cc-skills (ghostwriting and copywriting)Repo pass241MIT VectorBT Backtesting Skills (Marketcalls)Repo pass221– Swan GTM SkillsRepo pass181MIT Salesforce MCP Server (community)Repo pass165MIT Polymarket Python SDK (official)Pass138MIT Twenty CRM MCP Server (community)Repo pass106MIT Company OS Starter Kit (workflows.io)Pass92MIT PostHog AI plugin (official)Repo pass91– CCXT crypto market data MCPRepo pass82MIT Attio MCP Server (community)Repo pass69NOASSERTION Prediction Market Agent Tooling (Gnosis)Pass65LGPL-3.0 LeadMagic GTM skillsRepo pass64MIT Exa Agent Skills (official)Repo pass58MIT GTM Engineer Playbook (Othmane Khadri)Repo pass58MIT Stageshow (faux product UI skill, Benji Taylor)Repo pass51MIT GTM Engineer Starter (Jordan Crawford / Blueprint)Repo pass30MIT HubSpot Agent CLI Skills (official)Repo pass27Apache-2.0 Instantly MCP (community)Repo pass27MIT Zime call-review skillsRepo pass17MIT Pipedrive MCP Server (community, Nubiia)Repo pass12MIT Prosyo GTM SkillsRepo pass12MIT SignalForcePass10MIT HubSpot MCP plugins (official)Repo pass4Apache-2.0 Agent Coach daily check-in (Avatar 8)Pass–Avatar 8 Agent-native directory launch (Avatar 8)Pass–Avatar 8 Claude Code GTM repo layout (Avatar 8)Pass–Avatar 8 Content metric lock (Avatar 8)Pass–Avatar 8 Credential and entity separation checklist (Avatar 8)Pass–Avatar 8 Evidence-required claims rule (Avatar 8)Pass–Avatar 8 Forecast calibration scoring (Avatar 8)Pass–Avatar 8 Inbox triage protocol (Avatar 8)Pass–Avatar 8 Meeting prep brief (Avatar 8)Pass–Avatar 8 North-star definition lock (template) (Avatar 8)Pass–Avatar 8 North-star metric lock (Avatar 8)Pass–Avatar 8 Outbound first run (staging only) (Avatar 8)Pass–Avatar 8 Owner-blocker batching protocol (Avatar 8)Pass–Avatar 8 Paper-trading discipline (Avatar 8)Pass–Avatar 8 Publishing QA checklist (Avatar 8)Pass–Avatar 8 Release = verified live (checklist skill) (Avatar 8)Pass–Avatar 8 Venture Operator guardrails (Avatar 8)Pass–Avatar 8 receiving-code-reviewRepo pass301kMIT requesting-code-reviewRepo pass301kMIT test-driven-developmentRepo pass301kMIT code-reviewRepo pass295kMIT git-guardrails-claude-codeRepo pass295kMIT webapp-testingRepo pass182k– analyticsRepo pass55kMIT promptfoo-evalsRepo pass26kMIT ui-testRepo pass3.7k– google-analytics-mcpRepo pass3.5kApache-2.0 accessibilityRepo pass3.0kMIT best-practicesRepo pass3.0kMIT performanceRepo pass3.0kMIT seoRepo pass3.0kMIT web-quality-auditRepo pass3.0kMIT apify-content-analyticsRepo pass2.4k– firecrawl-dashboard-reportingRepo pass195ISC firecrawl-qaRepo pass195ISC mcpRepo pass29MITPopular but not yet scanned (12)
- classifier.dev (Jev) 425 GitHub stars, repo needs review
- golang-testing 282k GitHub stars, not yet security-scanned
- security-review 282k GitHub stars, not yet security-scanned
- frontend-code-review 159k GitHub stars, not yet security-scanned
- frontend-testing 159k GitHub stars, not yet security-scanned
- garrytan/gstack 137k GitHub stars, not yet security-scanned
- gstack 137k GitHub stars, not yet security-scanned
- browser-testing-with-devtools 108k GitHub stars, not yet security-scanned
- code-review-and-quality 108k GitHub stars, not yet security-scanned
- observability-and-instrumentation 108k GitHub stars, not yet security-scanned
- test-driven-development 108k GitHub stars, not yet security-scanned
- koala73/worldmonitor 89k GitHub stars, not yet security-scanned
Section: Evals & quality. For agents: full markdown table.