knowledge-base-health-check
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from multiple directories within a knowledge base, creating a vulnerability surface where malicious instructions embedded in those files could influence the agent's behavior during the audit.
- Ingestion points: The skill is instructed to read every article in
<KB>/Wiki/, every file in<KB>/RAW/, and every output in<KB>/Outputs/filed since the last health check (as specified in Phase 1 of SKILL.md). - Boundary markers: The instructions do not define any delimiters or provide 'ignore embedded instructions' warnings for the agent when it processes these external files.
- Capability inventory: The agent has the capability to write new report files to
<KB>/Outputs/and modify the<KB>/CHANGELOG.mdfile. - Sanitization: There are no explicit instructions to sanitize, escape, or validate the content retrieved from the knowledge base folders before the agent interpolates that data into its generated reports or log entries.
Audit Metadata