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 724 of 1,836
- View details
contrastive-unpaired-translationSkillDatabases & data
Lets your agent run image-to-image translation tasks using CUT, FastCUT, and SinCUT workflows with dataset prep.
Ready to connect★ 295
- View details
control-netSkillProductivity
Lets your agent route ControlNet source-checkout tasks to annotators via a Gradio interface.
Ready to connect★ 295
- View details
cookiecutter-data-scienceSkillDev tools
Lets your agent generate a standard data science project folder structure ready to work in.
Ready to connect★ 295
- View details
coremltoolsSkillAI & models
Lets your agent convert and work with machine learning models using Core ML Tools workflows.
Ready to connect★ 295
- View details
coronavirusSkillDev tools
Lets your agent curate and organize coronavirus research data following Folding@home repository conventions.
Ready to connect★ 295
- View details
creamSkillDev tools
Lets your agent route neural architecture search and model compression tasks to Cream-family vision models.
Ready to connect★ 295
- View details
crud-archive-runSkillMonitoring & ops
Lets your agent save all logs, traces, and outputs from a finished experiment run before cleanup deletes them.
Ready to connect★ 295
- View details
crud-otagent-supabaseSkillDatabases & data
Lets your agent look up model benchmark scores and build or refresh eval tables in a Supabase registry.
Ready to connect★ 295
- View details
crud-purge-below-gate-evalsSkillAI & models
Lets your agent delete incomplete evaluation job records that failed the quality gate.
Ready to connect★ 295
- View details
crud-purge-stale-eval-placeholdersSkillDatabases & data
Lets your agent delete old, dead placeholder rows from a job registry that never got scored.
Ready to connect★ 295
- View details
cs230-code-examplesSkillDev tools
Lets your agent fetch example CS230 course code for PyTorch or TensorFlow tasks.
Ready to connect★ 295
- View details
datagen-create-task-datasetSkillDatabases & data
Lets your agent build a fresh, snapshot-safe task dataset from an HF dataset, a codebase, or plain instructions.
Ready to connect★ 295
- View details
datagen-launchSkillProductivity
Lets your agent launch trace-generation jobs on HPC clusters from parquet task files.
Ready to connect★ 295
- View details
datagen-launch-irisSkillMonitoring & ops
Lets your agent launch, monitor, and clean up data-generation jobs on the Iris TPU cluster.
Ready to connect★ 295
- View details
datagen-reduce-dataset-snapshotsSkillDatabases & data
Lets your agent shrink the number of unique environment snapshots in a dataset below its cap by editing build logic.
Ready to connect★ 295
- View details
datagen-standard-launchSkillDatabases & data
Lets your agent run standard data generation jobs with vLLM or API completions across multiple nodes.
Ready to connect★ 295
- View details
db-benchmarkSkillDev tools
Lets your agent add or update repository read-function benchmark cases in the DB benchmark script.
Ready to connect★ 295
- View details
docs-deslopSkillFiles & storage
Lets your agent condense and reorganize docs and config comments using a fresh-context editor.
Ready to connect★ 295
- View details
eval-agentic-cleanupSkillAI & models
Lets your agent audit a finished agentic eval and recommend fixes, without writing anything.
Ready to connect★ 295
- View details
eval-agentic-launchSkillAI & models
Lets your agent launch and verify agentic model evaluation jobs on a cluster, including model selection and tunnel setup.
Ready to connect★ 295
- View details
eval-agentic-launch-irisSkillMonitoring & ops
Lets your agent launch, monitor, and clean up model evaluation jobs on Marin's Iris TPU or CoreWeave GPU clusters.
Ready to connect★ 295
- View details
eval-standard-cleanupSkillAI & models
Lets your agent collect finished math eval job results and record the scores in a SCORES.md tracker.
Ready to connect★ 295
- View details
eval-standard-launchSkillAI & models
Lets your agent launch a fixed math benchmark eval suite (MATH-500, AIME24, gsm8k) on finished model checkpoints.
Ready to connect★ 295
- View details
file-issueSkillFiles & storage
Lets your agent create a GitHub issue for a bug or improvement found during the session.
Ready to connect★ 295
- View details
monitor-cron-sweepSkillMonitoring & ops
Lets your agent gather and summarize job statuses across HPC clusters on a recurring schedule.
Ready to connect★ 295
- View details
monitor-cron-sweep-irisSkillMonitoring & ops
Lets your agent check the status of scheduled Iris TPU training and evaluation jobs every few hours.
Ready to connect★ 295
- View details
monitor-job-tablesSkillMonitoring & ops
Lets your agent format HPC job-status reports as tables grouped by job type with metrics and red-flags.
Ready to connect★ 295
- View details
monitor-restoreSkillMonitoring & ops
Lets your agent re-create a 3-hour monitoring loop that checks three compute clusters for issues.
Ready to connect★ 295
- View details
monitor-restore-irisSkillMonitoring & ops
Re-creates a scheduled job that checks Marin TPU cluster jobs every 3 hours and rescues stalled datagen runs.
Ready to connect★ 295
- View details
monitor-restore-unifiedSkillMonitoring & ops
Lets your agent re-register a recurring ops-monitor cron job for data-generation and training campaigns.
Ready to connect★ 295
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.