metabolomics-workbench-database
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied metabolite identifiers and search terms which are interpolated into API request URLs, creating a surface for potential indirect prompt injection.
- Ingestion points: User-provided search parameters and identifiers (such as PubChem CIDs and InChI Keys) found in SKILL.md and references/api_reference.md.
- Boundary markers: The documentation does not define delimiters or specific instructions for the agent to disregard instructions within data returned from the API.
- Capability inventory: The skill documentation lists network GET requests to the metabolomicsworkbench.org domain in SKILL.md and references/api_reference.md.
- Sanitization: No sanitization or validation logic for user-provided parameters is specified in the documentation.
Audit Metadata