pubchem-database
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS] (SAFE): The skill makes legitimate HTTP requests to 'pubchem.ncbi.nlm.nih.gov' using the requests library. These connections are for retrieving public scientific data and follow rate-limiting best practices.
- [COMMAND_EXECUTION] (SAFE): There are no instances of subprocess spawning or arbitrary command execution. All functions are focused on data querying and parsing via API.
- [DATA_EXFILTRATION] (SAFE): The skill does not access local sensitive files like SSH keys or environment variables. Data transmission is limited to the defined scientific API endpoints.
- [PROMPT_INJECTION] (SAFE): No patterns of prompt injection, role-play bypass, or system prompt extraction were found in the scripts or metadata.
- [DYNAMIC_EXECUTION] (SAFE): The scripts do not use eval, exec, or other dynamic code generation techniques. It relies on standard library-based API interaction.
Audit Metadata