kegg-database
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill facilitates network communication with
https://rest.kegg.jpto retrieve biological pathway and genomic data. This is a well-known scientific resource, and the network activity is strictly limited to fetching database entries as described in the skill's documentation. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external KEGG API responses, creating a surface for indirect prompt injection.
- Ingestion points: External data enters the agent context via the helper functions in
scripts/kegg_api.py(e.g.,kegg_get,kegg_list). - Boundary markers: No specific delimiters or "ignore instructions" warnings are used in the prompt interpolation logic.
- Capability inventory: The skill's scripts are restricted to read-only network operations; they do not contain file-system write capabilities, subprocess calls, or dynamic code execution functions.
- Sanitization: Search queries are URL-encoded using
urllib.parse.quote, but the content retrieved from the database is returned as raw text without further sanitization.
Audit Metadata