Playbooks · Coding tasks
Vet a skill before you install it
Written October 29, 2026
For a solo developer × vetting one skill before it goes into Claude Code × Claude Code, ending with a written decision — install, install with limits, or do not install — made before the skill runs anything.
Before you start
A skill is a folder of instructions: a SKILL.md file that tells the agent what to do, plus any scripts, hooks, and reference files it ships with. Installing one hands the agent new instructions and, depending on the skill, new tools and commands it may run. The description in the skill's frontmatter states what it is for; it does not prove what it does. This page vets one skill in one pass: a scanner run, a read of the skill as an attacker would read it, a provenance check, and a written decision.
Scanners pattern-match text. They do not run the skill, and they cannot tell a documented attack example from a live instruction. On one security-teaching skill — vbsec, the vbs-scan-security skill, 76 files — all three scanners below returned their maximum severity. vetskill 0.1.0 returned BLOCK at 100/100, with 16 high, 10 medium, and 7 low findings. skillvet 0.2.0 returned RED at 100/100. SkillSpector 2.12.0, in static-only mode, scored 100, CRITICAL, with a DO NOT INSTALL recommendation and 141 issues, at 81.6% coverage in a scan run on October 29, 2026. The findings were pattern matches on documented examples inside the skill's own rule files — exploit code, example hosts such as 169.254.169.254 and evil.com, and calls such as eval( shown as attacks to detect — not instructions the skill runs. A scanner verdict alone would block a teaching skill. The human read below is what separates examples from executable behaviour, and it is not optional.
What "done" looks like
At the end of one pass you have a short decision note for the skill: its name, the commit you looked at, the scanner verdict, what the attacker-read found, and one of three decisions — install, install with limits, or do not install. The skill is not installed until the note exists.
Steps
- Copy the skill to a throwaway folder; run nothing yet. Clone the repo or copy the skill folder somewhere outside your project, and note the exact commit. Do not run its install script, and do not open it in the agent yet. Every later step works on this copy, so a bad skill never touches a live project.
- Run one scanner over the copy. Any one of these three runs the same kind of static pass; pick one and run it from outside the skill folder:
vetskill scan <skill folder>— install withpip install git+https://github.com/bayuh85/vetskill. As of November 14, 2026, that repository address returns a 404, so the install does not currently work; run skillvet or SkillSpector below instead. Therevealcommand reports hidden characters; therulescommand lists the 21 rules it checks.npx -y skillvet <skill folder>— no install step; it prints a points table and a verdict.skillspector scan <skill folder> --no-llm— static-only mode, no model key needed. Install it first withuv tool install git+https://github.com/NVIDIA/skillspector.git(SkillSpector); without that step the scan command is not found. It checks 71 vulnerability patterns in 17 categories and scores 0–100.
- Read SKILL.md as an attacker. Read the whole file, not just the description. Check four things. First, the frontmatter: a skill with no
allowed-toolsfield and nohooksfield declares no tool scope, so the body text is the only statement of what it may run — on the vbsec skill, the frontmatter carried noallowed-toolsorhooksat all. Second, hidden text: run the scanner's reveal or hidden-character check, and look for instructions inside HTML comments. Third, every host, command, and file the skill names: a host the skill actually calls is different from a host named in an example, and a command the skill runs is different from a command it quotes to detect. Fourth, the install and helper scripts: read them without executing them, and note anything that downloads code, reads credentials, or phones home. - Check provenance before trusting the read. Confirm the repo is the one the skill claims to come from, that it carries a licence file, and that the commit you noted matches what you scanned. A skill copied into a chat message or a gist has no provenance to check: treat it as untrusted text, not as the author's release. Note the install path the README gives, and whether it runs a script you have now read.
- Write the decision note, then act on it. One entry: skill name, commit, scanner name and verdict, the attacker-read findings in one line each, and the decision. Install means the scanner findings are explained by the skill's own purpose and the read found nothing executable beyond it. Install with limits means the skill runs only on a copy of the project, or with the network-dependent and destructive options left off — on the vbsec skill, the live OSV.dev queries and the opt-in patch-applying mode are the options that change what the skill does. Do not install means any finding survives the read: a host the skill actually calls that it does not document, a script that downloads and runs code, or instructions hidden where a reader would not look.
- Re-scan after any update or delisting. When the skill changes, or when a scanner or registry removes it from a listing, repeat the pass on the new commit before reinstalling: run the scanner again, read the diff, and write a new decision note. A skill cleared once is cleared for that commit only.
Check the agent's work
- Check that the scanner ran on a copy, and that no install script ran before the decision note existed.
- Check that every high-severity scanner finding has a one-line explanation: example text, or executable behaviour. An unexplained high finding is an open finding, not a pass.
- Check that the frontmatter tool scope was read, including the case where it is absent. Absent scope is a finding, not a default pass.
- Check that the decision note names the commit. A decision without a commit does not survive the next update.
- Check that helper scripts were read, not run. A script that was executed during vetting has already done what vetting was meant to prevent.
What can go wrong
- Trusting the scanner verdict alone. Scanners flag example code and example hosts. On teaching material they return maximum severity for material that is meant to be read, not run. The verdict starts the read; it does not replace it.
- Trusting the description alone. The frontmatter description states intent. The body, the scripts, and the hooks state behaviour. Vet the behaviour.
- Running the install script first. An install script runs with your permissions, in your project. Reading it first costs minutes; running it first can cost the project.
- Skipping the re-scan. Skills update. A new commit can add a script, a hook, or a network call that the cleared commit did not have. The old decision note does not cover it.
Limits
- Static scanners read text. They do not execute the skill, so they cannot prove what a script does when it runs — only what its text says. Reading the scripts is part of the pass for that reason.
- Some scanner modes need a model API key. Without one, only the static scan runs, and a partial scan — SkillSpector reported 81.6% coverage on the skill above — is a partial inspection, not a complete one.
- This pass vets one skill at one commit. It says nothing about the next commit, a fork under the same name, or a skill with the same name from another repo.
Related playbooks
- Install and vet framework-vendor skills: Expo and Xcode
- Vet an Antigravity plugin before the one-click install: what the bundle can do, and what Google does not check
- Vet OpenBot before you install it
- Make sure your Claude Code skill actually loads when it should: test it in each agent before you rely on it
Sources
Sources: vbsec — skill scanned, vetskill, SkillSpector.
Back to: Coding playbooks hub · Also: Open an untrusted repo without running its code