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
55,257 results · page 764 of 1,842
- View details
convert-to-backtestSkillSearch
Lets your agent turn an optimiser or grid search notebook into a standalone backtest notebook.
Ready to connect★ 250
- View details
convert-to-optimiserSkillCommerce & finance
Lets your agent turn a trading backtesting notebook into a parameter optimisation notebook automatically.
Ready to connect★ 250
- View details
create-pr-descriptionSkillDev tools
Lets your agent draft a pull request description that follows the repo's PR template and current changes.
Ready to connect★ 250
- View details
create-variantSkillCommerce & finance
Lets your agent create a copy of a backtesting or optimiser notebook so you can try changes.
Ready to connect★ 250
- View details
file-test-bugSkillFiles & storage
Lets your agent create a GitHub issue reporting a failed integration test.
Ready to connect★ 250
- View details
flag-compiler-bugSkillProductivity
Lets your agent detect likely compiler bugs during coding and file de-duplicated bug reports.
Ready to connect★ 250
- View details
fleet-auditSkillDev tools
Lets your agent report which local git worktrees under tmp/ are safe to close or need a human look.
Ready to connect★ 250
- View details
hello-worldSkillDev tools
Lets your agent print a simple hello world message to test it works.
Ready to connect★ 250
- View details
investigate-integration-testSkillCloud & infra
Lets your agent investigate a failing integration test by downloading logs, analyzing the failure, and suggesting fixes.
Ready to connect★ 250
- View details
markdown-token-optimizerSkillFiles & storage
Analyzes your markdown files to trim extra words and reduce the tokens they use.
Ready to connect★ 250
- View details
orchestrate-parallel-workSkillProductivity
Splits a task into parallel sub-agent jobs with clear ownership and merges their results together.
Ready to connect★ 250
- View details
ralph-loopSkillAI & models
Lets your agent run an autonomous planning-implementation-review loop that turns a scoped task into reviewed commits and a draft PR.
Ready to connect★ 250
- View details
review-code-smellsSkillDev tools
Lets your agent review changed code for maintainability issues like duplication, dead code, and poor naming.
Ready to connect★ 250
- View details
review-docs-claimsSkillDev tools
Lets your agent review documentation for accuracy, prose quality, and leaked internal details.
Ready to connect★ 250
- View details
review-incan-source-qualitySkillDev tools
Lets your agent review Incan source files and report on readability, naming, and Python-like code quality.
Ready to connect★ 250
- View details
review-orchestrateSkillAI & models
Lets your agent run a multi-pass code review by spawning specialized reviewer roles and merging their findings into one report.
Ready to connect★ 250
- View details
review-rfcSkillDocs & knowledge
Lets your agent check an RFC document for formatting and content problems and fix them before submission.
Ready to connect★ 250
- View details
review-rust-proseSkillFiles & storage
Lets your agent review Rust code changes for comment quality and rustdoc accuracy.
Ready to connect★ 250
- View details
review-scopeSkillDev tools
Lets your agent check whether a code change actually matches the issue, RFC, or user request it was meant to address.
Ready to connect★ 250
- View details
review-test-styleSkillDev tools
Lets your agent review changed tests for style, unwrap/expect use, Result patterns, and missing coverage.
Ready to connect★ 250
- View details
run-variant-cycleSkillCommerce & finance
Lets your agent create and run a copy of a trading backtesting or optimiser notebook with changed settings.
Ready to connect★ 250
- View details
senseiSkillCloud & infra
Lets your agent iteratively refine a skill's frontmatter until routing tests pass.
Ready to connect★ 250
- View details
seobuild-onpageSkillSearch
Helps your agent write search-optimized web pages using live search results and AI-citation techniques.
Ready to connect★ 250
- View details
summarise-notebook-folderSkillCommerce & finance
Lets your agent read all experiment notebooks in a folder and write a summary README.
Ready to connect★ 250
- View details
vally-evalSkillCloud & infra
Lets your agent create, validate, and run Vally evaluation suites that test how agent skills perform.
Ready to connect★ 250
- View details
write-commit-messageSkillCommunication
Lets your agent draft commit messages that follow your workspace's naming convention.
Ready to connect★ 250
- View details
write-rfcSkillDocs & knowledge
Lets your agent draft a new RFC document for the Incan language project following the project's required structure.
Ready to connect★ 250
- View details
yeet-prSkillCommunication
Lets your agent push local repo changes to GitHub and open a review-ready or draft pull request.
Ready to connect★ 250
- View details
codedrobe-adapter-devSkillDev tools
Lets your agent add, inspect, repair, or validate CodeDrobe Core adapters for Chromium and Electron desktop apps.
Ready to connect★ 248
- View details
codedrobe-themeSkillAI & models
Lets your agent create, apply, and restore custom color themes for supported AI desktop apps.
Ready to connect★ 248
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.
54,826 of the 55,257 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.