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
55,543 results · page 110 of 1,852
- View details
mvpSkillDev tools
Guides your agent through building a minimal viable product, starting manual and staying scoped down.
Ready to connect★ 10k
- View details
processizeSkillDev tools
Turns your product idea into a manual, hands-on process you can start delivering before writing any code.
Ready to connect★ 10k
- View details
validate-ideaSkillDev tools
Lets your agent test whether a business idea is worth pursuing before you build anything.
Ready to connect★ 10k
- View details
ghaSkillAI & models
Lets your agent analyze GitHub Actions failures and find the root cause.
Ready to connect★ 10k
- View details
half-cloneSkillAI & models
Lets your agent copy the recent half of a conversation and drop older context to save tokens.
Ready to connect★ 10k
- View details
handoffSkillDocs & knowledge
Lets your agent write or update a handoff document so a fresh agent can pick up the work.
Ready to connect★ 10k
- View details
hn-summarizeSkillAI & models
Lets your agent fetch and summarize Hacker News top stories, articles, and comment threads.
Ready to connect★ 10k
- View details
private-github-searchSkillSearch
Lets your agent search the text of all your GitHub repos, including private ones, using a local copy.
Ready to connect★ 10k
- View details
quarter-cloneSkillAI & models
Lets your agent copy the most recent part of a conversation and drop older messages to save tokens.
Ready to connect★ 10k
- View details
Reddit Content FetcherSkillWeb & browsing
Lets your agent fetch Reddit posts and comments for research, even when Reddit blocks normal requests.
Ready to connect★ 10k
- View details
review-claudemdSkillFiles & storage
Lets your agent review recent conversations and suggest improvements to your CLAUDE.md files.
Ready to connect★ 10k
- View details
version-checkSkillAI & models
Tells your agent which Claude Code version to run or whether updating now is a good idea.
Ready to connect★ 10k
- View details
alpha-researchSkillSearch
Lets your agent search, read, ask questions about, and annotate academic papers.
Ready to connect★ 10k
- View details
alphafold2SkillMonitoring & ops
Lets your agent predict protein structures and check confidence metrics using AlphaFold2-style workflows.
Ready to connect★ 10k
- View details
boltzSkillProductivity
Lets your agent plan and run biomolecular structure predictions for proteins, complexes, ligands, and nucleic acids.
Ready to connect★ 10k
- View details
borzoiSkillMedia
Lets your agent predict how DNA variants and regulatory sequences affect gene expression using genomics models.
Ready to connect★ 10k
- View details
chai1SkillProductivity
Lets your agent run and review biomolecular structure predictions, including protein complexes with ligands and nucleic acids.
Ready to connect★ 10k
- View details
compute-env-setupSkillProductivity
Lets your agent set up a ready-to-run Python or R research computing environment before starting a heavy job.
Ready to connect★ 10k
- View details
contributingSkillProductivity
Guides your agent through making code, docs, or skill changes to the Feynman repo and preparing a pull request.
Ready to connect★ 10k
- View details
dockerSkillCloud & infra
Lets your agent run research code inside isolated Docker containers so experiments stay safe and separate from your system.
Ready to connect★ 10k
- View details
eli5SkillDev tools
Lets your agent explain papers and technical ideas in plain English with simple analogies and clear takeaways.
Ready to connect★ 10k
- View details
esmfold2SkillProductivity
Lets your agent predict protein structures from amino acid sequences using fast ESMFold-style folding workflows.
Ready to connect★ 10k
- View details
evo2SkillMedia
Lets your agent analyze DNA, RNA, or protein sequences by scoring likelihoods, predicting variant effects, or designing edits.
Ready to connect★ 10k
- View details
fair-esm2SkillProductivity
Lets your agent run protein language model analysis like embeddings, mutation scoring, and sequence clustering.
Ready to connect★ 10k
- View details
figure-composerSkillDatabases & data
Lets your agent plan and assemble a publication-grade multi-panel scientific figure from a claim or dataset.
Ready to connect★ 10k
- View details
figure-styleSkillDev tools
Lets your agent apply scientific plotting rules to make research figures clear, labeled, and reproducible.
Ready to connect★ 10k
- View details
indication-dossierSkillProductivity
Lets your agent build a sourced research report on a disease, drug target, or therapy with ranked evidence.
Ready to connect★ 10k
- View details
jobsSkillProductivity
Lets your agent check the status of running research jobs and scheduled follow-ups.
Ready to connect★ 10k
- View details
ligandmpnnSkillMedia
Lets your agent design protein sequences that fit around ligands and small molecules using LigandMPNN-style workflows.
Ready to connect★ 10k
- View details
managed-model-endpointsSkillAI & models
Lets your agent register and check AI model endpoints for research workflows, including health checks and routing.
Ready to connect★ 10k
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.
55,111 of the 55,543 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.