install-skill
Skill0 starsChanged 35 days agoMITView on GitHub
Symlinks a skill (or every skill) from this skills repo into ~/.claude/skills, so Claude Code can discover and run it, and a later `git pull` is the only update step there ever is. Use this whenever the user asks to install, add, set up, sync, or update any skill from this repo — even if they just say \"install the X skill\" or \"add a skill\" without naming this skill explicitly — and whenever someone is setting up a new machine and wants every skill at once (\"install all the skills\", \"set up my machine\"). Also the skill to reach for on the very first bootstrap: before install-skill itself is symlinked in, run its bundled script once directly, or ask Claude Code in plain language to do it. That's the one time this doesn't need to already be \"installed\" to work.
11 other projects use a version of this.
Adopted by
Most recently taken up by
- aaronjmars/miroshark-aeontheir own version, first seen Oct 21, 2026
- aeonfun/aeontheir own version, first seen Oct 20, 2026
- MiroShark/miroshark-aeontheir own version, first seen Oct 17, 2026
- Lilac-Labs/gini-agenttheir own version, first seen Oct 8, 2026
- Open-Curiosity/gini-agenttheir own version, first seen Oct 8, 2026
- aaronjmars/aeontheir own version, first seen Oct 8, 2026
Used with
Projects with it also use aeon-update, auto-workflow and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
aaronjmarsSkillaeon-update
Pull framework updates from the upstream Aeon repo into this instance - 3-way merges canon's new commits into a PR, never clobbering operator config.
Also in 4 projects that use this
aaronjmarsSkillauto-workflow
Two-mode aeon.yml workflow builder - analyze inspects URLs and emits a tiered, signal-verified skill-enablement plan plus an aeon.yml diff; enable flips slugs to enabled:true and opens a PR.
Also in 4 projects that use this
- +10 more from aaronjmars
Other versions11 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 11
- Lilac-Labs/gini-agent★ 1.9ksame name
- Open-Curiosity/gini-agent★ 1.9ksame name
- aaronjmars/aeon★ 789same name
- aeonfun/aeon★ 789same name
- roerohan/skills★ 29same name
- aaronjmars/miroshark-aeon★ 20same name
- MiroShark/miroshark-aeon★ 20same name
- schwepps/skills★ 16same name
- Mineru98/skills-store★ 13same name
- xiangu152/my_claude_skills★ 2same name
1 more not listed
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.
In this repositoryOne other file it gives its agents.
- learn-thisSkill
The file
We split this skill into the rules it gives an agent
3 rules. 0 are common, 1 is distinctive, and 2 are very specific to this project.
0
Common
Most of the 9 edited versions we read keep these.
1
Distinctive
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 62
Run it from anywhere inside the repo clone.
- Very specific: line 68
If ~/.claude/skills/<name> is a real file or directory (not a symlink), the script never touches it.
- Very specific: line 71
If it's a symlink pointing somewhere else (a stale or manual install), the script reports the conflict and stops.
“Common” is measured: kept by at least 6 in 10 of the 9 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
install-skill
Symlink a skill from this repo into the local Claude skills folder
(~/.claude/skills by default), so it's discoverable and runnable, while
the repo stays the one and only copy. There is nothing to reinstall after a
git pull: the symlink always resolves to whatever's currently in the repo.
Without this, getting a skill into ~/.claude/skills is a manual ln -s
that nobody remembers the exact form of, or a copy that quietly drifts from
the repo. This gives you one consistent phrase ("install the X skill") and
safety checks so it can't clobber anything.
The bootstrap problem (read this first)
A skill can only be invoked from a session that already discovers it, so
install-skill can't install itself the first time.
The fix is to run the bundled script once, by hand or by asking Claude Code directly in plain language, no skill required:
git clone https://github.com/aornelas/claude-skills.git
python3 claude-skills/skills/install-skill/scripts/install_skill.py install-skillOr tell Claude Code: "install the install-skill skill from this repo via a symlink, so git pull keeps it updated." Claude Code has local filesystem access and can run the same script itself.
Once install-skill is symlinked into ~/.claude/skills, everything after
that, including re-installing install-skill after an update, goes through
this skill for consistency and the safety checks below.
(Sandboxed or cloud sessions generally can't reach ~/.claude/skills, a
protected home-directory path. This only works from a session with direct
local filesystem access, i.e. Claude Code running on your machine.)
Workflow
- Make sure the repo is current.
git pullin the clone before installing anything, so the symlink points at the latest version from the start. Run the bundled script for what the user asked for:
# One named skill python skills/install-skill/scripts/install_skill.py <skill-name> # Every skill in the repo: first-time machine setup python skills/install-skill/scripts/install_skill.py --all # Just check status, changes nothing python skills/install-skill/scripts/install_skill.py --listRun it from anywhere inside the repo clone. The script locates the repo root from its own file path, so it doesn't matter what directory the shell is currently in.
- Read the result back plainly. The script reports one of five
outcomes per skill: installed, already installed (no-op), relinked,
conflict, or not found. Conflicts are the case worth explaining clearly:
- If
~/.claude/skills/<name>is a real file or directory (not a symlink), the script never touches it. Tell the user to remove or rename it manually if they want the repo-linked version instead. - If it's a symlink pointing somewhere else (a stale or manual install),
the script reports the conflict and stops. Re-run with
--forceto relink it, once you've confirmed with the user that's what they want.
- If
- Confirm and move on. Once installed, tell the user they're done. No
further steps, and
git pullalone keeps every symlinked skill current from then on. New skills only need installing once; a session started afterwards will pick them up.
Notes
- This uses
os.symlink, which needs a Unix-like filesystem (macOS/Linux). On Windows, symlinks need elevated permissions or Developer Mode, so flag this rather than assuming it'll just work. --target-dir(or theCLAUDE_SKILLS_DIRenv var) overrides the default~/.claude/skills, for anyone with a non-default Claude config location.- The script only ever creates symlinks or refuses. It never deletes or
overwrites a real file/directory, and never relinks an existing symlink
without
--force. Destructive-by-accident isn't a risk here by design. - Skills installed this way always point at the clone they were installed
from. If you move or delete the clone, re-run the install from the new
location with
--force.
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.

