chem-hazard-toxicity
Pass
Audited by Gen Agent Trust Hub on Aug 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's implementation is entirely consistent with its stated purpose of fetching scientific safety data. The code is well-structured and uses standard patterns for API interaction and file management without any suspicious side effects.
- [EXTERNAL_DOWNLOADS]: The script fetches data from
https://pubchem.ncbi.nlm.nih.gov, which is a well-known, authoritative, and trusted government resource for chemical information. This network activity is documented, expected, and necessary for the skill's primary function. - [DATA_EXFILTRATION]: There is no evidence of unauthorized data access or transmission. The script does not access sensitive local files (e.g., credentials, environment variables) and only communicates with the trusted PubChem domain to retrieve chemical metadata.
- [COMMAND_EXECUTION]: The skill uses standard command-line argument parsing and limits its file system operations to creating the designated output directory and writing the resulting JSON file. No arbitrary command execution, privilege escalation, or persistence mechanisms were detected.
Audit Metadata