Skip to content
Directory

Agent skills

Skills for AI coding agents, each with what it costs to load, how much of it is scripts and what it touches on your machine.

1,192 skills27 categories Measured from their files, refreshed weekly

How to read a skill

Tokens

What it costs to keep

Idle in context, in use when it runs, and up to with its linked files.

Model dependence

Scripts or instructions

Script-led, mixed or model-led: how much the model has to work out itself.

Touches

What it can reach

Shell, network, file writes, git and credentials, declared or detected.

Scores

Quality, security, repo health

Out of 100, each with what it is made of. Turn a card over to see them.

Showing 12 of 1192 skillsScript-led Mixed Agent Workflows Developer Tools Network Git Clear all
Mixed4,484 tokens in use

subagent-driven-development

322,961

Agent Workflows · Developer Tools

Runs an implementation plan with a fresh implementer subagent per task, a spec and quality review after each, and a final branch review.

Skill
17idle
4,484in use
10,486up to
Tokens, counted offlineView skill
Mixed2,220 tokens in use

systematic-debugging

322,961

Code Review And Quality · Developer Tools

A four-phase debugging process that finds the root cause before any fix is proposed.

Skill
21idle
2,220in use
5,201up to
Tokens, counted offlineView skill
Mixed2,375 tokens in use

mcp-builder

198,136

Developer Tools

Anthropic's skill for building MCP servers: research the API, implement tools in TypeScript or Python, test with the Inspector, then measure the server with an LLM evaluation harness.

Skill
61idle
2,375in use
29,563up to
Tokens, counted offlineView skill
Mixed7,253 tokens in use

skill-creator

198,136

Agent Workflows · Developer Tools

Anthropic's skill for making skills: Claude drafts a SKILL.md, runs test prompts with and without it, shows the results in a review viewer and tunes the description that triggers it.

Skill
64idle
7,253in use
18,660up to
Tokens, counted offlineView skill
Mixed833 tokens in use

webapp-testing

198,136

Code Review And Quality · Developer Tools

Anthropic's skill for testing local web apps: Claude writes native Python Playwright scripts, starts the dev servers with a helper, then inspects screenshots, the DOM and console logs.

Skill
35idle
833in use
833up to
Tokens, counted offlineView skill
Mixed2,794 tokens in use

dbs-install-skill

11,518

Agent Workflows

Links one skill, or a folder of skills, into every agent installed on the machine with a single script, and cleans up duplicate entries.

Skill
87idle
2,794in use
2,794up to
Tokens, counted offlineView skill
Mixed2,106 tokens in use

shadcn-ui

11,108

Developer Tools

Guides agents through shadcn/ui work: finding components, installing them with the CLI, theming with CSS variables and extending them with cva.

Skill
31idle
2,106in use
14,070up to
Tokens, counted offlineView skill
Mixed1,255 tokens in use

goal-prompt

8,320

Agent Workflows

Drafts one-line /goal commands for Claude Code and Codex that have a checkable end state, a stated check, invariants and a stop bound.

Skill
67idle
1,255in use
1,255up to
Tokens, counted offlineView skill
Script-led1,093 tokens in use

Dual Axis Skill Reviewer

7,471

Agent Workflows

Scores agent skills on two axes, deterministic code checks and an LLM review, then merges them into a weighted score with improvement items.

Skill
96idle
1,093in use
1,703up to
Tokens, counted offlineView skill
Mixed592 tokens in use

skill-designer

7,471

Agent Workflows

Builds a prompt from a JSON skill idea that tells Claude CLI to create a full skill folder with SKILL.md, references, scripts and tests.

Skill
48idle
592in use
2,174up to
Tokens, counted offlineView skill
Script-led9,245 tokens in use

Task Observer

3,987

Agent Workflows

A meta-skill that logs corrections and repeated workflows during agent sessions, then stages evidenced updates to your skills for your approval.

Skill
197idle
9,245in use
145,052up to
Tokens, counted offlineView skill
Mixed2,660 tokens in use

Claude Agents

687

Agent Workflows

Drive headless Claude Code (claude -p) from another agent as a peer, a builder or a fresh reviewer, through a wrapper that keeps sessions and modes.

Skill
40idle
2,660in use
5,351up to
Tokens, counted offlineView skill

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers