skillguard
Skill112 starsChanged 3 months agoMITView on GitHub
Security scanner for AgentSkill packages. Scan skills for credential theft, code injection, prompt manipulation, data exfiltration, and evasion techniques before installing them. Use when evaluating skills from ClawHub or any untrusted source.
- Reads credentials
2 other projects use a version of this.
Adopted by
Most recently taken up by
- LLMSecurity/skillguardtheir own version, first seen Feb 25, 2026
- Muhammad-Qasim-Munir/skillguardtheir own version, first seen Jan 30, 2026
Other versions2 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 2
- LLMSecurity/skillguard★ 8same name
- Muhammad-Qasim-Munir/skillguard★ 4same 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.
In this repository14 other files it gives its agents.
- AGENTS.mdAGENTS.md
- cron-packsSkill
- memory-upgradeSkill
- mission-controlSkill
- everclawSkill
- agent-chatSkill
- bagmanSkill
- night-shiftSkill
- pii-guardSkill
- prompt-guardSkill
- relationshipsSkill
- three-shiftsSkill
- xmtp-comms-guardSkill
- three-shiftsSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 9
When asked to check, audit, or scan a skill for security, use SkillGuard.
- Very specific: line 30
Use --untrusted to force external mode on an internal skill:
- Distinctive: line 69
Default: full text report
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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
SkillGuard — Agent Security Scanner
When asked to check, audit, or scan a skill for security, use SkillGuard.
Trust Model (v0.3)
SkillGuard distinguishes between internal (trusted) and external (untrusted) skills:
- Internal skills — Skills under
~/.openclaw/workspace/skills/or~/.openclaw/workspace/claw-repos/. These are skills we authored ourselves. - External skills — Everything else (ClawHub downloads, third-party packages, etc.)
Why it matters
Internal skills legitimately use patterns like exec(), process.env.API_KEY, fetch(), and writeFile() — that's infrastructure code doing its job. Flagging those as "suspicious" creates noise that buries real issues.
For internal skills, SkillGuard uses vulnerability-focused scanning:
- ✅ Still flags: hardcoded secrets, reverse shells, pickle deserialization, unsafe YAML, actual code obfuscation, prompt injection in operational code
- ❌ Suppresses: "uses exec", "reads env vars", "makes HTTP requests", "writes files", "references private keys" (in wallet tools), "uses sudo" (in setup scripts), behavioral compound signatures
For external skills, full paranoid threat-model scanning applies — every pattern is treated as potentially malicious.
Overriding trust
Use --untrusted to force external mode on an internal skill:
node src/cli.js scan /path/to/skill --untrustedCommands
Scan a local skill directory
node /home/claw/.openclaw/workspace/skillguard/src/cli.js scan <path>Scan with compact output (for chat)
node /home/claw/.openclaw/workspace/skillguard/src/cli.js scan <path> --compactCheck text for prompt injection
node /home/claw/.openclaw/workspace/skillguard/src/cli.js check "<text>"Batch scan multiple skills
node /home/claw/.openclaw/workspace/skillguard/src/cli.js batch <directory>Scan a ClawHub skill by slug
node /home/claw/.openclaw/workspace/skillguard/src/cli.js scan-hub <slug>Score Interpretation
- 80-100 ✅ LOW risk — safe to install
- 50-79 ⚠️ MEDIUM — review findings before installing
- 20-49 🟠 HIGH — significant security concerns
- 0-19 🔴 CRITICAL — do NOT install without manual review
Output Formats
- Default: full text report
--compact: chat-friendly summary--json: machine-readable full report--quiet: score and verdict only
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.