1,000 open-source agent skills
The best agent skills on GitHub
Agent skills are folders with a SKILL.md file that teach AI coding agents such as Claude Code, Codex and Cursor how to do one task well. appsgit lists a curated, growing collection of 1,000 developer-focused open-source skills from skills.sh and GitHub, each with install commands, trigger phrases and live GitHub stats.
- Agent skills
- 1,000
- Official
- 195
- Categories
- 7
- Updated
- Oct 26, 2026
Trending now
Agent skills gaining stars fastest
Ranked by GitHub stars gained between two of appsgit's daily star snapshots 3 to 14 days apart, as close to 7 days as the data allows, scaled to a 7-day rate so repositories with different snapshot gaps compare fairly. A repository is held for review and left out of the ranking (it stays listed) when its gain scaled to 7 days is over 50,000 stars, when more than half of its stars arrived inside the window (repositories with 5,000 stars or more), or when it leads with at least 1,000 stars and more than 10 times the next repository's gain. A human releases a held repository after review.
Security Audit
cloudflare/security-audit-skill
Security guidance and vulnerability review for codebases, APIs, services, CLI tools, libraries, and daemons.
Maintained · last commit Sep 14, 2026
+16,175 stars in the last 7 days
OfficialSecurityJavaScriptPonytail
DietrichGebert/ponytail
Lazy senior dev mode: the smallest change that fully solves the task, and a reply a busy human understands in one read.
Active · last commit Oct 18, 2026
+9,088 stars in the last 7 days
Docs & writingJavaScriptPonytail Audit
DietrichGebert/ponytail
Quality audit of a whole repo: bugs, security holes, what breaks under real load, risky code without tests, slow paths, and what to delete, merge or split.
Active · last commit Oct 18, 2026
+9,088 stars in the last 7 days
SecurityJavaScriptComputer Use
stablyai/orca
Drives the GUI of a visible local app window through orca computer: accessibility tree, clicks, typing, menus, dialogs, and screenshots in native apps and external browser windows (Chrome, Edge…
Active · last commit Oct 26, 2026
+9,006 stars in the last 7 days
OfficialDesign & frontendTypeScriptOrca Chat Visuals
stablyai/orca
Show a chart, diagram, table, report or mockup inline in this Orca chat as an HTML page.
Active · last commit Oct 26, 2026
+9,006 stars in the last 7 days
OfficialData & analysisTypeScriptCode Review
mattpocock/skills
Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes: Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match…
Active · last commit Oct 19, 2026
+6,873 stars in the last 7 days
CodingShell
Most starred
Top 12 agent skills
Ranked by GitHub stars. Every page has install instructions, required keys and live repository stats.
Systematic Debugging
obra/superpowers
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
Active · last commit Oct 21, 2026
Testing & QAShellWriting Plans
obra/superpowers
Use when you have a spec or requirements for a multi-step task, before touching code.
Active · last commit Oct 21, 2026
Docs & writingShellRequesting Code Review
obra/superpowers
Use when completing tasks, implementing major features, or before merging to verify work meets requirements.
Active · last commit Oct 21, 2026
CodingShellTest Driven Development
obra/superpowers
Use when implementing any feature or bugfix, before writing implementation code.
Active · last commit Oct 21, 2026
Testing & QAShellVerification Before Completion
obra/superpowers
Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims…
Active · last commit Oct 21, 2026
Testing & QAShellReceiving Code Review
obra/superpowers
Use when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not…
Active · last commit Oct 21, 2026
CodingShellWriting Skills
obra/superpowers
Use when creating new skills, editing existing skills, or verifying skills work before deployment.
Active · last commit Oct 21, 2026
Docs & writingShellFinishing A Development Branch
obra/superpowers
Use when implementation is complete, all tests pass, and you need to decide how to integrate the work.
Active · last commit Oct 21, 2026
Testing & QAShellDiagnosing Superpowers
obra/superpowers
Use when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so…
Active · last commit Oct 21, 2026
Testing & QAShellImprove Codebase Architecture
mattpocock/skills
Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.
Active · last commit Oct 19, 2026
CodingShellTDD
mattpocock/skills
Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.
Active · last commit Oct 19, 2026
Testing & QAShellSetup Matt Pocock Skills
mattpocock/skills
Configure this repo for the engineering skills: set up its issue tracker, triage label vocabulary, and domain doc layout.
Active · last commit Oct 19, 2026
CodingShell
Browse by category
Agent skills by category
- Coding132 agent skillsRequesting Code Review, Receiving Code Review, Improve Codebase Architecture
- Testing & QA153 agent skillsSystematic Debugging, Test Driven Development, Verification Before Completion
- Docs & writing146 agent skillsWriting Plans, Writing Skills, Writing Beats
- Data & analysis136 agent skillsMigration, Orca Chat Visuals, Dating Web
- DevOps & cloud78 agent skillsCaveman Setup, Paperclip Board, Childhood
- Design & frontend208 agent skillsPrototype, Codebase Design, Frontend Design
- Security147 agent skillsWizard, Ponytail Audit, Next Bundle Optimizer
What is an agent skill?
A skill is a folder containing a SKILL.md file: YAML frontmatter with a name and a description, followed by instructions, and optionally scripts or reference files. Agents load only the short description until a task matches it, then read the full instructions. The format is an open standard (agentskills.io) supported by Claude, Codex, Cursor, Gemini CLI, GitHub Copilot and others.
How do I add a skill to Claude?
In Claude Code, copy the skill folder (the one containing SKILL.md) to ~/.claude/skills/<skill-name>/ for all your projects, or to .claude/skills/<skill-name>/ inside one project. If the repository is a plugin marketplace, run /plugin marketplace add <owner>/<repo> and then /plugin install <plugin>@<marketplace>. In the Claude apps (claude.ai and desktop), zip the skill folder and upload it under Customize > Skills > + > Upload a skill; code execution must be turned on. Source: code.claude.com, code.claude.com, support.claude.com
How do I add a skill to ChatGPT or Codex?
Codex reads skills from .agents/skills/<skill-name>/ in your repository (and its parent folders) and from ~/.agents/skills/ (or $CODEX_HOME/skills) for every project, so copy the skill folder there. Inside Codex, the built-in $skill-installer skill can install skills from OpenAI's catalog. Standalone skills also work in the ChatGPT desktop app; skills bundled in plugins work on ChatGPT web, desktop and mobile. Source: learn.chatgpt.com
How do I add a skill to Cursor?
Cursor loads agent skills from .cursor/skills/ or .agents/skills/ in a project and from ~/.cursor/skills/ or ~/.agents/skills/ for all projects. It also reads Claude's .claude/skills/ folders, so a skill installed for Claude Code is picked up by Cursor too. Source: cursor.com
What does Official mean?
Official means the skill is published by the vendor itself: Anthropic's own repositories, organizations listed as official on skills.sh, or GitHub organizations with a verified domain. 195 skills on appsgit are official.
How are skills chosen?
Skills come from skills.sh, the GitHub topics claude-skills, agent-skills, claude-code-skills and skills, and the major awesome lists. A skill is listed only if its SKILL.md has a valid name and a description of 80+ characters, it has a license (the repository's or its own), the repository has a commit in the last six months, and the repository has 100+ stars or the skill is listed on skills.sh. This first batch is curated for developers (coding, testing, DevOps, data, security, frontend and technical docs) and capped at about 1,000 skills, at most 25 per repository; more categories are coming.
Are agent skills safe?
Skills are instructions plus optional scripts that your agent may run, so only install skills from sources you trust. Read the SKILL.md and any bundled scripts first, and prefer official or widely used skills.