openclaw-guardian@1.0.1
CAUTION· install onlyPublisher claims: "A security layer plugin for OpenClaw that intercepts dangerous tool calls (exec, write, edit) through two-tier regex blacklist rules and LLM-based intent ver..." Observed in sandbox: no network, touches nothing outside the project, writes no memory.
Coverage
Nothing executable in the bundle ran; the trace covers the install step only.- Tested
- every file read, file written and host contacted by what ran
- Not tested
- use of the skill by an agent
Is [email protected] safe to install?
KillerSkills ran it in the KillerSkills honeypot sandbox (1 scenario; nothing executable in the bundle ran; the trace covers the install step only) on 2026-11-18 and scored it 68/100 (Caution). It is not known to be malicious, but it has behaviour or code a security policy should decide on before install. Notable findings: Destructive command against a filesystem root or device; Access to sensitive file paths; Host or command assembled from base64 fragments. Summary: no network, touches nothing outside the project, writes no memory.
Behavior trace
Honeypot run · 0 s · 1 scenarios · 0 deviationsStatic findings
User said: "echo 'rm -rf /' 这个命令很危险"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.- Writes to `/etc/passwd`, `/etc/shadow`, `/etc/sudoers`Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.- `rm -rf` on safe paths, `sudo`, `chmod 777`, `chown root`Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.- Pipe attacks: `curl | sh`, `wget | bash`, `base64 -d | sh`Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.| Warning | 1 | ~1-2s | `rm -rf /tmp/cache`, `chmod 777`, `sudo apt` |Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.- Warning: dotfiles, `/etc/` configs, `.env` files, `authorized_keys`Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.- Pipe attacks: `curl | sh`, `wget | bash`, `base64 -d | sh`Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.Tool: exec, Command: rm -rf /var/log/oldFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.{ pattern: /xargs\s+.*\bchmod\b/, reason: "xargs chmod (indirect permission change)" },Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.| Warning | 1 | ~1-2s | `rm -rf /tmp/cache`, `chmod 777`, `sudo apt` |Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.- Pipe attacks: `curl | sh`, `wget | bash`, `base64 -d | sh`Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.Tool: exec, Command: rm -rf /data/old-backupFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.User said: "echo 'rm -rf /' 这个命令很危险"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.Tool: exec, Command: rm -rf /tmp/cacheFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.Tool: exec, Command: rm -rf /Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.- Chain attacks: download + `chmod +x` + executeFix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.Tool: exec, Command: rm -rf /home/user/projectFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.Tool: exec, Command: sudo systemctl restart nginxFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.| Warning | 1 | ~1-2s | `rm -rf /tmp/cache`, `chmod 777`, `sudo apt` |Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.reason: "download + chmod +x chain (download and execute)",Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.{ pattern: /chmod\s+[47]77\b/, reason: "world-writable permission (chmod 777)" },Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.reason: "chown on system path",Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.// Force killFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.{ pattern: /kill\s+-9\s+/, reason: "force kill process (SIGKILL)" },Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.Cost to the agent
Tokens estimated at four characters each. Skills load their instruction text when triggered; the figure is what that costs.
Known vulnerabilities
Vulnerability advisories (OSV) cover npm and PyPI packages; this registry has no advisory database.
What would make this Verified
12 points to 80 · weights: behavior 45%, static 25%, prompt 15%, publisher 15%- Destructive command against a filesystem root or device≈ +9 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
- Unclaimed publisher≈ +8 on the totalClaim fatcatmaofei: a verified publisher gets +50 on publisher trust. It is free and usually instant with GitHub sign-in.
- Obfuscated code≈ +5 on the totalRemove base64-assembled strings, eval of decoded text and minified blobs. Reviewers and scanners both treat unreadable code as hostile.
- Access to sensitive file paths≈ +4 on the totalStop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
- Host or command assembled from base64 fragments≈ +4 on the totalPut hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
- Destructive file system command detected≈ +4 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.