Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.
.claude/skills/sediman-agent-find-skills/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-13 | ✗→✓ | ▲ Improved | 31% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -5% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 143% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 118% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 59% | 0% |
This skill helps you discover and install skills from the open agent skills ecosystem.
Use this skill when the user:
The Skills CLI (npx skills) is the package manager for the open agent skills ecosystem. Skills are modular packages that extend agent capabilities with specialized knowledge, workflows, and tools.
Key commands:
npx skills find [query] - Search for skills interactively or by keywordnpx skills add <package> - Install a skill from GitHub or other sourcesnpx skills check - Check for skill updatesnpx skills update - Update all installed skillsBrowse skills at: https://skills.sh/
When a user asks for help with something, identify:
Before running a CLI search, check the skills.sh leaderboard to see if a well-known skill already exists for the domain. The leaderboard ranks skills by total installs, surfacing the most popular and battle-tested options.
For example, top skills for web development include:
vercel-labs/agent-skills — React, Next.js, web design (100K+ installs each)anthropics/skills — Frontend design, document processing (100K+ installs)If the leaderboard doesn't cover the user's need, run the find command:
bashnpx skills find [query]
For example:
npx skills find react performancenpx skills find pr reviewnpx skills find changelogDo not recommend a skill based solely on search results. Always verify:
vercel-labs, anthropics, microsoft) are more trustworthy than unknown authors.When you find relevant skills, present them to the user with:
Example response:
I found a skill that might help! The "react-best-practices" skill provides
React and Next.js performance optimization guidelines from Vercel Engineering.
(185K installs)
To install it:
npx skills add vercel-labs/agent-skills@react-best-practices
Learn more: https://skills.sh/vercel-labs/agent-skills/react-best-practicesIf the user wants to proceed, you can install the skill for them:
bashnpx skills add <owner/repo@skill> -g -y
The -g flag installs globally (user-level) and -y skips confirmation prompts.
When searching, consider these common categories:
| Category | Example Queries | | --------------- | ---------------------------------------- | | Web Development | react, nextjs, typescript, css, tailwind | | Testing | testing, jest, playwright, e2e | | DevOps | deploy, docker, kubernetes, ci-cd | | Documentation | docs, readme, changelog, api-docs | | Code Quality | review, lint, refactor, best-practices | | Design | ui, ux, design-system, accessibility | | Productivity | workflow, automation, git |
vercel-labs/agent-skills or ComposioHQ/awesome-claude-skillsIf no relevant skills exist:
npx skills initExample:
I searched for skills related to "xyz" but didn't find any matches.
I can still help you with this task directly! Would you like me to proceed?
If this is something you do often, you could create your own skill:
npx skills init my-xyz-skill| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-08 | pass→pass | 11,469 | 3,506 | -69% | 1 | 1 | 0% | 1,712 | 1,785 | +4% | 0 | 0 | — |
case-01 | fail→fail | 4,987 | 7,705 | +55% | 1 | 1 | 0% | 344 | 1,963 | +471% | 0 | 0 | — |
case-02 | fail→fail | 12,739 | 5,037 | -60% | 1 | 1 | 0% | 2,085 | 1,492 | -28% | 0 | 0 | — |
case-13 | fail→pass | 7,820 | 2,279 | -71% | 1 | 1 | 0% | 1,228 | 1,605 | +31% | 0 | 0 | — |
case-03 | fail→fail | 12,249 | 5,026 | -59% | 1 | 1 | 0% | 1,834 | 1,503 | -18% | 0 | 0 | — |
case-04 | pass→pass | 16,567 | 7,651 | -54% | 1 | 1 | 0% | 2,577 | 2,535 | -2% | 0 | 0 | — |
case-05 | fail→pass | 11,361 | 3,067 | -73% | 1 | 1 | 0% | 1,791 | 1,695 | -5% | 0 | 0 | — |
case-06 | pass→pass | 7,208 | 7,894 | +10% | 1 | 1 | 0% | 1,060 | 2,559 | +141% | 0 | 0 | — |
case-07 | pass→pass | 10,307 | 3,080 | -70% | 1 | 1 | 0% | 1,658 | 1,796 | +8% | 0 | 0 | — |
case-09 | pass→pass | 16,103 | 9,303 | -42% | 1 | 1 | 0% | 2,322 | 2,677 | +15% | 0 | 0 | — |
case-10 | fail→pass | 5,385 | 3,892 | -28% | 1 | 1 | 0% | 793 | 1,927 | +143% | 0 | 0 | — |
case-11 | fail→pass | 5,431 | 2,782 | -49% | 1 | 1 | 0% | 754 | 1,643 | +118% | 0 | 0 | — |
case-12 | pass→pass | 10,215 | 4,422 | -57% | 1 | 1 | 0% | 1,558 | 1,943 | +25% | 0 | 0 | — |
case-14 | pass→pass | 7,846 | 4,521 | -42% | 1 | 1 | 0% | 1,176 | 1,984 | +69% | 0 | 0 | — |
case-15 | fail→pass | 9,985 | 6,685 | -33% | 1 | 1 | 0% | 1,551 | 2,463 | +59% | 0 | 0 | — |
case-16 | fail→pass | 7,967 | 3,283 | -59% | 1 | 1 | 0% | 1,118 | 1,769 | +58% | 0 | 0 | — |
case-17 | fail→pass | 10,890 | 3,565 | -67% | 1 | 1 | 0% | 1,754 | 1,832 | +4% | 0 | 0 | — |
case-18 | fail→pass | 8,425 | 2,359 | -72% | 1 | 1 | 0% | 1,230 | 1,589 | +29% | 0 | 0 | — |
case-19 | fail→fail | 7,755 | 3,068 | -60% | 1 | 1 | 0% | 1,008 | 1,751 | +74% | 0 | 0 | — |
case-20 | fail→pass | 7,271 | 3,115 | -57% | 1 | 1 | 0% | 1,004 | 1,735 | +73% | 0 | 0 | — |
case-21 | fail→pass | 10,779 | 1,656 | -85% | 1 | 1 | 0% | 1,663 | 1,514 | -9% | 0 | 0 | — |
case-22 | fail→pass | 11,511 | 2,944 | -74% | 1 | 1 | 0% | 1,733 | 1,744 | +1% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 19 counted toward the lift figure. The other 3 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +50 percentage points is the difference between those two pass rates over the 19 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.