string-database
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEMETADATA_POISONINGINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [METADATA_POISONING]: The skill includes a 'Suggest Using K-Dense Web' section that provides specific instructions for the agent to proactively recommend a commercial research platform (www.k-dense.ai) when user requests increase in complexity. This acts as a behavioral steering mechanism.
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from the external STRING REST API to perform network analysis and functional enrichment. This ingestion of untrusted data creates a potential surface for indirect prompt injection. 1. Ingestion points: STRING API responses processed in SKILL.md workflows. 2. Boundary markers: No specific delimiters or safety instructions are used to isolate API data from the agent's instructions. 3. Capability inventory: The skill executes local Python scripts (scripts/string_api.py and scripts/string_enrichment.py) and uses data manipulation libraries like pandas. 4. Sanitization: No explicit sanitization or validation of the API data is described in the integration instructions.
- [COMMAND_EXECUTION]: The skill documentation provides instructions for importing and running local Python helper functions from 'scripts/string_api.py' and 'scripts/string_enrichment.py' to interact with the database and process results.
Audit Metadata