folklore-variant-evidence
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill documentation includes a
curlexample for making JSON-RPC calls to a remote service. This is used to demonstrate how to programmatically interact with the Folklore MCP endpoint. - [EXTERNAL_DOWNLOADS]: The skill connects to
api.helena.bioto fetch gene-disease associations and scientific evidence. This external network communication is necessary for the skill's operation. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from external scientific sources which could theoretically contain adversarial content.
- Ingestion points: The tools
search_variant_evidence,search_variant_literature, andsearch_literature_corpusfetch content from external bioinformatics and literature databases. - Boundary markers: The skill contains specific instructions to 'Enforce the input boundary' to prevent the inclusion of sensitive patient data.
- Capability inventory: The skill performs network-based data retrieval to query scientific evidence.
- Sanitization: The instructions require the agent to filter out all patient-specific identifiers, phenotypes, and clinical records from its queries.
Audit Metadata