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 675 of 1,836
- View details
Fetch Agent CardSkillAI & models
Lets your agent fetch an agent card JSON file from a given URL to inspect what that agent offers.
Ready to connect★ 336
- View details
framing-research-questionsSkillDatabases & data
Guides your agent to frame clear research questions before starting any data analysis or investigation.
Ready to connect★ 335
- View details
implementation-notesSkillDocs & knowledge
Lets your agent keep a running implementation-notes.md logging plan deviations and edge cases during builds.
Ready to connect★ 335
- View details
implementation-planSkillDev tools
Lets your agent write implementation plans that put likely-to-change decisions first and routine work last.
Ready to connect★ 335
- View details
investigating-anomalous-resultsSkillAI & models
Guides your agent through a structured investigation when an experiment result looks surprising or wrong.
Ready to connect★ 335
- View details
pitch-packagerSkillDocs & knowledge
Lets your agent turn finished work like specs or prototypes into a single summary document for reviewers.
Ready to connect★ 335
- View details
preregistering-analysisSkillDev tools
Lets your agent lock in predictions and decision rules before running a confirmatory analysis or looking at outcome data.
Ready to connect★ 335
- View details
progressive-disclosureSkillFiles & storage
Lets your agent split a large instruction file into a small entry file that loads extra details only when needed.
Ready to connect★ 335
- View details
receiving-critical-reviewSkillDev tools
Guides your agent to verify and question critical feedback on analyses or manuscripts before making changes.
Ready to connect★ 335
- View details
reference-huntSkillDev tools
Lets your agent study an existing codebase or site you point to and copy its behavior into new code.
Ready to connect★ 335
- View details
reporting-and-archiving-findingsSkillDev tools
Helps your agent decide how to report completed analyses and archive the work so it can be reproduced.
Ready to connect★ 335
- View details
requesting-red-team-reviewSkillDev tools
Has a skeptical reviewer attack your agent's conclusions before it reports them as findings.
Ready to connect★ 335
- View details
setting-up-reproducible-analysisSkillDev tools
Sets up an isolated, reproducible analysis workspace with pinned environment, fixed seeds, and locked raw data.
Ready to connect★ 335
- View details
subagent-driven-analysisSkillAI & models
Lets your agent run a pre-planned analysis by breaking it into independent steps and delegating each to subagents.
Ready to connect★ 335
- View details
surveying-prior-workSkillDev tools
Lets your agent search past research to check methods, novelty, and effect sizes before starting an analysis.
Ready to connect★ 335
- View details
test-blindspotsSkillDev tools
Finds important behaviors your passing tests don't actually cover using focused exploratory checks.
Ready to connect★ 335
- View details
using-science-superpowersSkillDev tools
Sets up your agent to check for and use available skills before responding to any request.
Ready to connect★ 335
- View details
verifying-results-before-claimingSkillDev tools
Makes your agent run an analysis fresh and read the actual output before claiming any result.
Ready to connect★ 335
- View details
writing-science-skillsSkillCloud & infra
Lets your agent create, edit, and verify its own science-related skills before deployment.
Ready to connect★ 335
- View details
action-first-outputSkillAI & models
Formats the agent's replies to lead with the action, then numbered steps and a state summary.
Ready to connect★ 335
- View details
additive-bias-defenseSkillProductivity
Challenges your agent to question whether every code addition is truly necessary during reviews.
Ready to connect★ 335
- View details
agent-expenditureSkillAI & models
Lets your agent track token usage per agent and spot waste in parallel multi-agent runs.
Ready to connect★ 335
- View details
Ai Hygiene AuditSkillAI & models
Lets your agent scan a codebase for quality problems common in AI-generated code.
★ 335
- View details
Analyze SkillSkillFiles & storage
Analyzes skill files for complexity and suggests how to split them into smaller, simpler parts.
★ 335
- View details
Arch InitSkillAI & models
Lets your agent set up new projects with architecture-aware templates matched to the target domain.
★ 335
- View details
architecture-aware-initSkillDocs & knowledge
Lets your agent research and pick a software architecture paradigm with documented justification before building a project.
Ready to connect★ 335
- View details
architecture-paradigm-client-serverSkillAI & models
Guides your agent to design apps using a client-server architecture with centralized backends and trust boundaries.
Ready to connect★ 335
- View details
architecture-paradigm-cqrs-esSkillAI & models
Helps your agent design systems that separate reads from writes and keep a full history of state changes.
Ready to connect★ 335
- View details
architecture-paradigm-domain-drivenSkillAI & models
Guides your agent to design software around business rules using domain-driven design.
Ready to connect★ 335
- View details
architecture-paradigm-event-drivenSkillCommunication
Teaches your agent to design systems where components communicate through events instead of direct calls.
Ready to connect★ 335
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.