openclaw-parallels-smoke
Pass
Audited by Gen Agent Trust Hub on May 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains operational guidelines for software testing and interpretation. No malicious code, exfiltration attempts, or obfuscation were detected.
- [CREDENTIALS_SAFE]: While the skill references the use of API keys (OpenAI, Anthropic, and Discord), it includes explicit security instructions to avoid printing profile contents or environment dumps, ensuring secrets are handled safely through the guest execution environment.
- [COMMAND_EXECUTION]: Command execution is restricted to standard development tools (npm, pnpm), system utilities (timeout, curl, apt-get), and virtual machine management tools (prlctl). These are necessary for the skill's stated purpose of running automated tests.
- [EXTERNAL_DOWNLOADS]: External operations are limited to trusted package registries (npm) and platform-specific installers required for bootstrapping the test environment, which is standard for a development/QA skill.
Audit Metadata