tooluniverse-chemical-safety
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill integrates with established scientific and regulatory platforms, including the FDA, PubChem (NIH), ChEMBL (EMBL-EBI), and DrugBank. All external interactions are conducted through structured tool calls to these recognized services.
- [SAFE]: No indicators of prompt injection, role-play bypasses, or instructions to ignore safety guidelines were found in the instructional content.
- [SAFE]: There are no hardcoded credentials, sensitive file access attempts, or unauthorized data exfiltration patterns. The network operations are limited to the intended scientific data retrieval.
- [SAFE]: The skill does not employ obfuscation techniques. SMILES strings present in the documentation and test scripts are standard chemical notation and are used correctly within the skill's logic.
- [SAFE]: The test script and quick start guides follow standard development practices, utilizing a specialized library ('tooluniverse') which is consistent with the vendor's naming conventions.
Audit Metadata