find-a-skill
Skill145 starsChanged 8 days agoMITView on GitHub
Find and install agent skills from SkillMD (skillmd.com), the open registry of over a million Agent Skills. Use when the user asks for a skill for something, asks whether a skill exists for a task, or needs know-how none of your installed skills cover, such as a file format, framework, API or workflow you have no instructions for.
3 other projects use a version of this.
Adopted by
Most recently taken up by
- skillmds/skillmdword for word, first seen Nov 15, 2026
- ubik-dsh/deepseek-harness-skillstheir own version, first seen Oct 11, 2026
- ubik-dsh/deepseek-modstheir own version, first seen Sep 26, 2026
Other versions3 projects use a version of this.
Word for word · 1
- skillmds/skillmd★ 145100% same
Edited · 0
None found.
Wrote their own version · 2
- ubik-dsh/deepseek-harness-skills★ 0same name
- ubik-dsh/deepseek-mods★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 16
Search for what the user wants done ("fill PDF forms", "write Playwright tests"), not for a tool name you guess at.
- Distinctive: line 24
You can tell where it came from: Prefer skills marked ✓ verified, with a license and a source repository you can link to.
- Distinctive: line 26
If nothing passes, say so and do the task without a skill rather than install a weak match.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Find a skill
SkillMD (https://skillmd.com) indexes over a million Agent Skills and checks each one for safety. When a task needs instructions you don't have, look there before working from scratch.
1. Search
- With the SkillMD MCP server connected, call
skillmd_searchwith a short description of the task. - Otherwise run
npx get-skill search "<the task in a few words>" --limit 10, or open https://skillmd.com/search?q=<the task>.
Search for what the user wants done ("fill PDF forms", "write Playwright tests"), not for a tool name you guess at.
2. Vet the best matches before suggesting any
Take the three to five most relevant results and look each one up with npx get-skill info <owner/name> (or skillmd_get). Drop any that fails one of these checks:
- It does this job. The description covers the task itself, not just its topic. A narrow skill that does exactly the job beats a popular general one.
- It asks only for what the job needs.
securitylists what the skill can do:docs_onlymeans instructions only;executes_scripts,network_callsandreads_secretsare fine when the task calls for them. A writing guide that reads secrets, or a formatter that calls the network, is a reason to skip it. - You can tell where it came from. Prefer skills marked ✓ verified, with a license and a source repository you can link to.
If nothing passes, say so and do the task without a skill rather than install a weak match.
3. Let the user choose
Show the user the one to three that passed, one line each: what it does, why it fits, and anything it can do beyond reading instructions (scripts, network, secrets). Install only the one they pick:
npx get-skill add <owner/name>Then carry on with the task using it.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

