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
54,221 results · page 106 of 1,808
- View details
arxivSkillSearch
Lets your agent search arXiv, download papers, and summarize them into a local library.
Ready to connect★ 16k
- View details
auto-paper-improvement-loopSkillAI & models
Lets your agent automatically revise a generated paper by reviewing it, applying fixes, and recompiling over two rounds.
Ready to connect★ 16k
- View details
auto-review-loopSkillAI & models
Lets your agent run repeated rounds of self-review on its research, applying fixes each time until approved.
Ready to connect★ 16k
- View details
auto-review-loop-llmSkillAI & models
Lets your agent run an autonomous review loop on research using any OpenAI-compatible LLM API.
Ready to connect★ 16k
- View details
auto-review-loop-minimaxSkillAI & models
Lets your agent run a multi-round external review of its work using the MiniMax AI model.
Ready to connect★ 16k
- View details
citation-auditSkillAI & models
Lets your agent verify that every citation in a paper is real, correctly attributed, and used in a supporting context.
Ready to connect★ 16k
- View details
claims-draftingSkillAI & models
Lets your agent draft patent claims for an invention.
Ready to connect★ 16k
- View details
comm-lit-reviewSkillProductivity
Lets your agent search communications research literature on wireless, networking, and satellite topics.
Ready to connect★ 16k
- View details
deepxivSkillSearch
Lets your agent search and read sections of open-access academic papers.
Ready to connect★ 16k
- View details
dse-loopSkillMedia
Lets your agent run architecture simulations, analyze results, and tune parameters repeatedly until it finds the best configuration.
Ready to connect★ 16k
- View details
embodiment-descriptionSkillAI & models
Lets your agent write detailed embodiment descriptions for patent specifications.
Ready to connect★ 16k
- View details
experiment-auditSkillAI & models
Lets your agent audit AI experiment results for fake data, scoring tricks, and overstated claims.
Ready to connect★ 16k
- View details
experiment-bridgeSkillCloud & infra
Lets your agent turn an experiment plan into running code: implements it, deploys to GPU, and collects results.
Ready to connect★ 16k
- View details
experiment-planSkillAI & models
Turns a research proposal into a detailed experiment roadmap with ablations, evaluation steps, and compute budget.
Ready to connect★ 16k
- View details
experiment-queueSkillAI & models
Lets your agent queue and run many machine learning experiments in batches over SSH, retrying those that run out of memory.
Ready to connect★ 16k
- View details
feishu-notifySkillCommunication
Lets your agent send notification messages to Feishu or Lark.
Ready to connect★ 16k
- View details
figure-descriptionSkillAI & models
Lets your agent turn patent figures into formal drawing descriptions with reference numerals.
Ready to connect★ 16k
- View details
figure-specSkillAI & models
Lets your agent generate editable SVG architecture, workflow, and pipeline diagrams from structured JSON.
Ready to connect★ 16k
- View details
formula-derivationSkillDocs & knowledge
Lets your agent organize equations and assumptions into a clear step-by-step formula derivation.
Ready to connect★ 16k
- View details
gemini-searchSkillSearch
Lets your agent search research papers using Gemini to find literature beyond standard academic indexes.
Ready to connect★ 16k
- View details
grant-proposalSkillAI & models
Lets your agent draft structured grant proposals for funders like NSF, ERC, and KAKENHI from your research ideas.
Ready to connect★ 16k
- View details
idea-creatorSkillAI & models
Lets your agent generate and rank research ideas from a broad topic you give it.
Ready to connect★ 16k
- View details
idea-discoverySkillAI & models
Guides your agent through a step-by-step workflow to turn a broad research direction into tested ideas.
Ready to connect★ 16k
- View details
idea-discovery-robotSkillAI & models
Lets your agent survey robotics research, generate and critique new research ideas with novelty checks.
Ready to connect★ 16k
- View details
integrity-forensicsSkillAI & models
Lets your agent run a forensic integrity sweep on a research paper and turn the findings into a policy verdict.
Ready to connect★ 16k
- View details
interview-cheatsheetSkillAI & models
Lets your agent generate a long-form Chinese ML/LLM interview cheat sheet with formulas, PyTorch code, and question lists.
Ready to connect★ 16k
- View details
invention-structuringSkillAI & models
Lets your agent turn a rough invention idea into a formal, patent-ready invention disclosure.
Ready to connect★ 16k
- View details
jurisdiction-formatSkillDocs & knowledge
Lets your agent format a patent application to meet the filing requirements of a specific jurisdiction like CN, US, or EP.
Ready to connect★ 16k
- View details
kill-argumentSkillAI & models
Lets your agent stress-test a paper draft by writing a hostile rejection memo and then rebutting it point by point.
Ready to connect★ 16k
- View details
mermaid-diagramSkillAI & models
Lets your agent create Mermaid diagrams like flowcharts, sequence diagrams, and Gantt charts from your requirements.
Ready to connect★ 16k
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.
53,789 of the 54,221 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.