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

54,220 results · page 435 of 1,808

  • common-skillsSkillAI & models

    Teaches your agent best practices for managing HuggingFace models and GGUF files in LlamaFarm.

    Ready to connect★ 838

    github.com/llama-farm/llamafarm837 stars

    View details
  • config-skillsSkillAI & models

    Teaches your agent how to build and validate LlamaFarm configuration files using Pydantic, JSONSchema, and YAML.

    Ready to connect★ 838

    github.com/llama-farm/llamafarm837 stars

    View details
  • designer-skillsSkillDev tools

    Gives your agent design patterns and guidance for building frontends with React 18, TanStack Query, TailwindCSS, and Radix UI.

    Ready to connect★ 838

    github.com/llama-farm/llamafarm837 stars

    View details
  • electron-skillsSkillSecurity

    Teaches your agent Electron app patterns for building desktop software with processes, IPC, security, and packaging.

    Ready to connect★ 838

    github.com/llama-farm/llamafarm837 stars

    View details
  • generate-subsystem-skillsSkillDev tools

    Lets your agent create specialized skills and checklists for each subsystem in a monorepo.

    Ready to connect★ 838

    github.com/llama-farm/llamafarm837 stars

    View details
  • go-skillsSkillDev tools

    Shared Go best practices for LlamaFarm CLI. Covers idiomatic patterns, error handling, and testing.

    Ready to connect★ 838

    github.com/llama-farm/llamafarm837 stars

    View details
  • python-skillsSkillSecurity

    Gives your agent Python coding best practices for patterns, async, typing, testing, error handling, and security.

    Ready to connect★ 838

    github.com/llama-farm/llamafarm837 stars

    View details
  • rag-skillsSkillAI & models

    Gives your agent best-practice guidance for building retrieval-augmented generation with LlamaIndex, ChromaDB, and Celery.

    Ready to connect★ 838

    github.com/llama-farm/llamafarm837 stars

    View details
  • react-skillsSkillDev tools

    Gives your agent React 18 coding patterns, including components, hooks, TanStack Query, and testing.

    Ready to connect★ 838

    github.com/llama-farm/llamafarm837 stars

    View details
  • runtime-skillsSkillDocs & knowledge

    Gives your agent best-practice guidance for running PyTorch models with Transformers and serving them via FastAPI.

    Ready to connect★ 838

    github.com/llama-farm/llamafarm837 stars

    View details
  • server-skillsSkillDev tools

    Gives your agent best-practice guidance for writing FastAPI, Celery, and Pydantic server code.

    Ready to connect★ 838

    github.com/llama-farm/llamafarm837 stars

    View details
  • temp-filesSkillFiles & storage

    Teaches your agent to create reports, logs, and progress files in the system temp directory instead of cluttering the repository.

    Ready to connect★ 838

    github.com/llama-farm/llamafarm837 stars

    View details
  • typescript-skillsSkillDev tools

    Lets your agent follow shared TypeScript coding conventions for the Designer and Electron subsystems.

    Ready to connect★ 838

    github.com/llama-farm/llamafarm837 stars

    View details
  • align-humanSkillDev tools

    Checks whether your automatic judge's scores agree with human labels and flags where it is biased.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • auto-arenaSkillProductivity

    Lets your agent compare multiple AI models by generating test queries, judging responses, and ranking win rates.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • bib-verifySkillFiles & storage

    Lets your agent check a BibTeX file for fake or wrong references by cross-checking each entry against CrossRef, arXiv, and DBLP.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • bootstrapSkillDev tools

    Lets your agent build a starter evaluation for your app from scratch, generating a v0 grader, test inputs, and a calibration roadmap.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • claude-authenticitySkillAI & models

    Lets your agent check whether an API endpoint serves genuine Claude and uncover injected system prompts.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • eval-designSkillDatabases & data

    Helps your agent design evaluation datasets with stratified test cases and adversarial examples for testing AI systems.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • eval-reportSkillDev tools

    Lets your agent combine results from multiple evaluation runs into one report with maturity scores, trends, and prioritized fixes.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • meta-evalSkillAI & models

    Helps your agent figure out how to evaluate LLM app quality by asking a few questions and recommending the right workflow.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • metric-designSkillDatabases & data

    Lets your agent design evaluation metrics, write judge prompts, and build automated grading pipelines.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • openjudgeSkillAI & models

    Lets your agent build and run pipelines that score and grade outputs from language models.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • paper-reviewSkillFiles & storage

    Lets your agent review academic papers for correctness, quality, and novelty from PDF or LaTeX files.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • prompt-regressionSkillAI & models

    Lets your agent compare two prompt versions head-to-head and report which is better with statistical confidence.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • rag-evalSkillAI & models

    Lets your agent evaluate a RAG system's quality by checking whether problems come from retrieval or from generation.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • redteamSkillSecurity

    Lets your agent run safety and security tests on your LLM app, checking for jailbreaks, prompt injection, and PII leaks.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • ref-hallucination-arenaSkillSearch

    Lets your agent benchmark how accurately AI models cite real academic papers by checking references against scholarly databases.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • rl-rewardSkillProductivity

    Helps your agent design reward signals for reinforcement learning training, choosing scoring strategies and combining scores.

    Ready to connect★ 836

    github.com/agentscope-ai/openjudge837 stars

    View details
  • agentcookie-installSkillMonitoring & ops

    Installs a tool that syncs your Chrome cookies from your Mac to another machine so your agent can share your browser logins.

    Ready to connect★ 833

    github.com/mvanhorn/agentcookie814 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.

53,788 of the 54,220 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