Skills.

Give your AI a better way to work.

A skill is a set of written instructions that teaches an AI how to do one job the way it should be done: review a pull request, plan a migration, write the release notes.

Install one here and it travels with your account into Claude, Claude Code, Cursor and every other client you sign in with.

Filter by category

55,055 results · page 725 of 1,836

  • php-developmentSkillSecurity

    Gives your agent expert guidance for writing PHP 8+ code that follows quality, security, and testing practices.

    Ready to connect★ 295

    github.com/cecilapp/cecil295 stars

    View details
  • repo-skills-routerSkillDatabases & data

    Routes your agent's AI, data, and software questions to the right managed repository skills based on what the request mentions.

    Ready to connect★ 295

    github.com/vectorspacelab/arex-skill85 stars

    View details
  • rl-agentic-job-cleanupSkillAI & models

    Lets your agent preserve and publish the best RL training checkpoint after a training job finishes or is cancelled.

    Ready to connect★ 295

    github.com/open-thoughts/openthoughts-agent286 stars

    View details
  • rl-agentic-launch-irisSkillCloud & infra

    Lets your agent launch or relaunch SkyRL GRPO reinforcement learning training jobs on Marin's Iris GPU cluster.

    Ready to connect★ 295

    github.com/open-thoughts/openthoughts-agent286 stars

    View details
  • rl-agentic-launch-jupiterSkillAI & models

    Lets your agent launch and relaunch agentic reinforcement learning training jobs on the Jupiter supercomputer.

    Ready to connect★ 295

    github.com/open-thoughts/openthoughts-agent286 stars

    View details
  • rl-job-health-deep-diveSkillMonitoring & ops

    Lets your agent probe the health of an RL training job and return a KILL / NO-KILL / ERROR recommendation.

    Ready to connect★ 295

    github.com/open-thoughts/openthoughts-agent286 stars

    View details
  • rl-standard-job-cleanupSkillAI & models

    Lets your agent clean up and publish a finished standard SkyRL reinforcement learning training job on Leonardo.

    Ready to connect★ 295

    github.com/open-thoughts/openthoughts-agent286 stars

    View details
  • rl-standard-launch-leonardoSkillDatabases & data

    Lets your agent launch and manage standard reinforcement learning training runs for math and reasoning models on the Leonardo supercomputer.

    Ready to connect★ 295

    github.com/open-thoughts/openthoughts-agent286 stars

    View details
  • search-paramsSkillSearch

    Lets your agent read and write URL query strings for filters, pagination, and shareable link state.

    Ready to connect★ 295

    github.com/sendou-ink/sendou.ink295 stars

    View details
  • sendou-code-reviewSkillSecurity

    Lets your agent run a multi-angle code review of your current changes and write findings to a file.

    Ready to connect★ 295

    github.com/sendou-ink/sendou.ink295 stars

    View details
  • serve-model-vibe-testSkillWeb & browsing

    Lets your agent set up a public web endpoint where people can test a machine learning model.

    Ready to connect★ 295

    github.com/open-thoughts/openthoughts-agent286 stars

    View details
  • sft-cleanup-hf-onlySkillDatabases & data

    Cleans up a finished HF-only SFT model run by uploading to Hugging Face without database registration.

    Ready to connect★ 295

    github.com/open-thoughts/openthoughts-agent286 stars

    View details
  • sft-job-cleanupSkillDatabases & data

    Lets your agent finish and clean up a completed LLaMA-Factory SFT training job by uploading the model, registering it, and freeing disk space.

    Ready to connect★ 295

    github.com/open-thoughts/openthoughts-agent286 stars

    View details
  • sft-launchSkillAI & models

    Lets your agent launch supervised fine-tuning jobs on major HPC clusters using LLaMA-Factory or axolotl.

    Ready to connect★ 295

    github.com/open-thoughts/openthoughts-agent286 stars

    View details
  • supervisor-initSkillAI & models

    Sets up a supervisor role that acts as the single point of contact between the user and its subagents and scheduled tasks.

    Ready to connect★ 295

    github.com/open-thoughts/openthoughts-agent286 stars

    View details
  • utils-cleanup-stale-sandboxesSkillAI & models

    Lets your agent delete old Daytona sandboxes across all three orgs in one pass.

    Ready to connect★ 295

    github.com/open-thoughts/openthoughts-agent286 stars

    View details
  • utils-fix-permissionsSkillFiles & storage

    Lets your agent fix file permissions on a cluster directory so teammates can read your shared files.

    Ready to connect★ 295

    github.com/open-thoughts/openthoughts-agent286 stars

    View details
  • utils-reclaim-stale-snapshotsSkillMedia

    Lets your agent delete idle Daytona snapshots org-wide to free space under the 60-snapshot cap.

    Ready to connect★ 295

    github.com/open-thoughts/openthoughts-agent286 stars

    View details
  • accessibility-a11ySkillAI & models

    Guides your agent to build accessible web pages with semantic HTML, keyboard navigation, ARIA labels, and WCAG contrast.

    Ready to connect★ 292

    github.com/blackbelttechnology/pi-agent-dashboard294 stars

    View details
  • adopt-stylesSkillDev tools

    Lets your agent apply breaking changes from fundamental-styles to Angular components.

    Ready to connect★ 294

    github.com/sap/fundamental-ngx294 stars

    View details
  • anti-slop-frontendSkillAI & models

    Lets your agent check AI-generated frontend code against a checklist that flags common AI design clichés.

    Ready to connect★ 292

    github.com/blackbelttechnology/pi-agent-dashboard294 stars

    View details
  • apple-toolsSkillSearch

    Lets your agent read and search your Apple Calendar, Contacts, Reminders, Messages, location, Maps, and Weather on macOS.

    Ready to connect★ 292

    github.com/blackbelttechnology/pi-agent-dashboard294 stars

    View details
  • bpmn-package-explorerSkillFiles & storage

    Lets your agent turn a written business process description into BPMN diagrams and decision tables, then validate and render them.

    Ready to connect★ 292

    github.com/blackbelttechnology/pi-agent-dashboard294 stars

    View details
  • build-formSkillDev tools

    Lets your agent build reactive forms with validation and error handling using Fundamental NGX form components.

    Ready to connect★ 294

    github.com/sap/fundamental-ngx294 stars

    View details
  • build-page-layoutSkillDev tools

    Lets your agent build a web page layout with a collapsing header, content area, footer, and tabs.

    Ready to connect★ 294

    github.com/sap/fundamental-ngx294 stars

    View details
  • build-tableSkillDev tools

    Lets your agent build a data table with sorting, filtering, pagination, and row selection using Fundamental NGX.

    Ready to connect★ 294

    github.com/sap/fundamental-ngx294 stars

    View details
  • canvas-webappSkillMedia

    Lets your agent display a bundled React or Vite web app on a dashboard canvas and fix blank or erroring canvas pages.

    Ready to connect★ 292

    github.com/blackbelttechnology/pi-agent-dashboard294 stars

    View details
  • ci-troubleshootSkillFiles & storage

    Lets your agent diagnose failed GitHub Actions CI runs, read failure logs, and know which workflow does what.

    Ready to connect★ 292

    github.com/blackbelttechnology/pi-agent-dashboard294 stars

    View details
  • component-architectureSkillMedia

    Gives your agent patterns for building reusable UI components like cards, forms, and layouts with consistent props.

    Ready to connect★ 292

    github.com/blackbelttechnology/pi-agent-dashboard294 stars

    View details
  • dashboard-plugin-scaffoldSkillAI & models

    Lets your agent create a new dashboard plugin project or add dashboard plugin features to an existing project.

    Ready to connect★ 292

    github.com/blackbelttechnology/pi-agent-dashboard294 stars

    View details

What is a skill?

A skill is plain text, usually a SKILL.md file and the scripts it refers to, written for an AI rather than for a person. It carries the steps, the house rules and the examples a good answer needs, so you stop pasting the same briefing into every new chat.

54,624 of the 55,055 skills listed here can be served through ahel today, and they come from public repositories. Each one has its own page with the instructions themselves on it, so you can read what a skill will tell your AI to do before you install it.

Install one and every AI you use gets it

Installing a skill adds it to your gateway and turns it on in the same step. Claude Code surfaces it as a slash command; any client can read the full instructions with the skill_read tool.

Nothing is copied into a project folder. The instructions are served from your account, so the same skill is there in every AI you connect, and turning it off removes it from all of them at once.

See how to connect your AI