openclaw-skill-eval@1.1.1
VERIFIED· install onlyPublisher claims: "Skill evaluation framework. Use when: testing trigger rate, quality compare (with/without skill), or model comparison. Runs via sessions_spawn + sessions_his..." Observed in sandbox: no network, touches nothing outside the project, writes no memory.
Coverage
Bundled scripts ran with no arguments; no agent task scenario for this bundle.- Tested
- 8 of 15 bundled scripts run with no arguments (the rest were not run: per-bundle cap), with decoy credentials in the environment · 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 (2 scenarios; bundled scripts ran with no arguments; no agent task scenario for this bundle) on 2026-11-20 and scored it 85/100 (Verified). In what ran, nothing we could detect went wrong: no undeclared network egress, no credential reads, no persistent instruction writes. What it does with a real credential was not observed; the static findings and the publisher carry the rest of the score. Summary: no network, touches nothing outside the project, writes no memory.
Behavior trace
Honeypot run · 1 s · 2 scenarios · 0 deviationsStatic findings
help="Show detailed output per eval (default: summary only)"Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.help="Show detailed output per eval (default: summary only)"Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.help="Show detailed output per eval (default: summary only)"Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.help="Show detailed output per eval (default: summary only)"Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.├── trigger_results.json ← analyze_triggers outputFix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.- "parallel": create two versions per eval (fresh + with-context)Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.result = subprocess.run(cmd, cwd=Path(__file__).parent)Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out."""Process a single eval (fetch histories, extract transcripts, save files).Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.help="Show detailed output per eval (default: summary only)"Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.result = subprocess.run(cmd, cwd=Path(__file__).parent)Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.├── quality_results.json ← analyze_quality outputFix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.result = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.help="Number of runs per eval (min 3 for meaningful stats). Default: 5"Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.# Port still in use after kill attempt — find a free oneFix: 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.