open-ontologies
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides a local Python utility (
scripts/preflight.py) and instructs the agent to execute it to verify the integrity of external source code. It also references an external binary (open-ontologies) for semantic planning and reasoning tasks, contingent on user approval.- [SAFE]: The providedpreflight.pyscript implements high-standard security practices to prevent common vulnerabilities. It uses secure POSIX file descriptor traversal (os.O_NOFOLLOWanddir_fd) to prevent symlink-based race conditions and path traversal. Additionally, it enforces a strict blocklist for sensitive files (e.g., SSH keys, cloud credentials, environment files, and private keys) and limits file reads to a maximum of 8 MiB per file. The skill design explicitly separates source verification from runtime execution to ensure a secure operating environment.
Audit Metadata