bkit-rules
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill classifies tasks and triggers agents based on keywords in user input.
- Ingestion points: User task descriptions and intent strings processed in the Task Classification section.
- Boundary markers: Absent; the skill does not define specific delimiters to isolate user-supplied content from the classification logic.
- Capability inventory: The skill's capabilities are restricted to the Read, Glob, and Grep tools.
- Sanitization: Absent; keywords are matched against raw user input without escaping or filtering.
Audit Metadata