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
54,680 results · page 607 of 1,823
- View details
agentsop-dspySkillAI & models
Guides your agent to build and optimize LLM prompt programs using the DSPy framework's compile-and-evaluate workflow.
Ready to connect★ 421
- View details
agentsop-framework-selectionSkillAI & models
Helps your agent pick the right LLM, agent, or RAG framework by walking a neutral decision tree at project kickoff.
Ready to connect★ 421
- View details
agentsop-http-tool-wrappingSkillAI & models
Guides your coding agent through turning an existing REST, GraphQL, or RPC API into a well-designed tool it can call.
Ready to connect★ 421
- View details
agentsop-hybrid-retrievalSkillAI & models
Guides your agent to add keyword search alongside embedding search so exact codes and names in your documents are found reliably.
Ready to connect★ 421
- View details
agentsop-idempotent-ingestionSkillProductivity
Lets your agent build and debug document-indexing pipelines that re-run without duplicating or leaving stale data.
Ready to connect★ 421
- View details
agentsop-langgraphSkillMedia
Guides your coding agent through designing, debugging, and running stateful LangGraph agent workflows.
Ready to connect★ 421
- View details
agentsop-llamaindexSkillDocs & knowledge
Guides your agent to build, debug, and evaluate RAG pipelines over private documents using LlamaIndex patterns.
Ready to connect★ 421
- View details
agentsop-llm-artifact-versioningSkillCloud & infra
Lets your agent version a whole LLM app bundle—prompts, model pins, configs—so deploys are reproducible and rollback is simple.
Ready to connect★ 421
- View details
agentsop-llm-engine-selectionSkillCloud & infra
Guides your agent to pick the right self-hosted LLM serving engine (vLLM, SGLang, llama.cpp, etc.) for your hardware and workload.
Ready to connect★ 421
- View details
agentsop-llm-tool-idempotencySkillAI & models
Guides your agent through designing tool calls so retried calls do not repeat actions like double-sending.
Ready to connect★ 421
- View details
agentsop-map-reduce-fanoutSkillAI & models
Guides your agent to split list-processing tasks into parallel model calls and combine the results safely.
Ready to connect★ 421
- View details
agentsop-metric-designSkillSearch
Helps your agent design reliable scoring metrics for evaluating LLM outputs, breaking them into smaller yes/no checks instead of one fuzzy judgment.
Ready to connect★ 421
- View details
agentsop-module-shape-selectionSkillAI & models
Helps your agent pick the right reasoning style (direct answer, chain of thought, tool loop, or code) before writing a prompt.
Ready to connect★ 421
- View details
agentsop-multi-tenant-ragSkillSecurity
Guides your agent to build and review retrieval systems that keep each tenant's data isolated from other users.
Ready to connect★ 421
- View details
agentsop-multiscale-chunkingSkillDocs & knowledge
Guides your agent to split long documents into small chunks for precise search while returning bigger surrounding context for better answers.
Ready to connect★ 421
- View details
agentsop-observability-setupSkillMonitoring & ops
Helps your agent pick the right LM tracing backend and turn on one-line logging so every AI call is recorded.
Ready to connect★ 421
- View details
agentsop-output-format-by-modelSkillProductivity
Guides your agent to choose the right output format for a task based on the model and who will consume the result.
Ready to connect★ 421
- View details
agentsop-per-model-artifactsSkillDatabases & data
Guides your agent through managing compiled prompts so they stay reliable when models, datasets, or framework versions change.
Ready to connect★ 421
- View details
agentsop-prompt-compilationSkillMonitoring & ops
Checks whether your prompt is ready for auto-optimization before you spend money running an optimizer.
Ready to connect★ 421
- View details
agentsop-prompt-history-inspectSkillAI & models
Lets your agent dump the exact prompt a language model call actually received to debug unexpected outputs.
Ready to connect★ 421
- View details
agentsop-query-routingSkillDatabases & data
Helps your agent send each query to the right index, tool, or answering engine before searching.
Ready to connect★ 421
- View details
agentsop-regression-gateSkillMonitoring & ops
Lets your agent build a held-out eval set and gate prompt or model changes in CI so regressions fail the build.
Ready to connect★ 421
- View details
agentsop-repo-mapSkillFiles & storage
Lets your agent build a compact map of a codebase's key classes and functions so it can find the right files to edit.
Ready to connect★ 421
- View details
agentsop-repo-state-gatingSkillAI & models
Lets your agent classify a codebase as new, large, familiar, or SDK, then pick the right coding tools and autonomy level.
Ready to connect★ 421
- View details
agentsop-reranker-stageSkillDocs & knowledge
Guides your agent to add and tune a reranker stage that improves which documents a RAG pipeline surfaces first.
Ready to connect★ 421
- View details
agentsop-selfhost-decisionSkillAI & models
Helps your agent decide whether to self-host an LLM or use a managed cloud API using volume and compliance criteria.
Ready to connect★ 421
- View details
agentsop-session-state-hygieneSkillProductivity
Helps your agent decide when to clear or keep its coding session context and spot stale-history bias.
Ready to connect★ 421
- View details
agentsop-signature-designSkillMedia
Helps your agent decide when to turn a long hand-written prompt into a typed DSPy Signature and how to name its fields.
Ready to connect★ 421
- View details
agentsop-state-reducerSkillAI & models
Guides your coding agent to declare reducers on LangGraph state keys so parallel graph runs merge updates instead of crashing.
Ready to connect★ 421
- View details
agentsop-streaming-outputSkillAI & models
Guides your agent to stream long LLM and tool run output from the backend to users.
Ready to connect★ 421
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.
54,249 of the 54,680 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.