tool-selection
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill takes an arbitrary operation description as input to determine tool routing, creating a surface for indirect prompt injection attacks.
- Ingestion points: The
[operation]argument in the/tools:selectcommand defined inreferences/select.md. - Boundary markers: Absent; there are no instructions for the agent to use delimiters or ignore embedded commands within the input string.
- Capability inventory: The skill coordinates with tools including
execute_sketched_edit(code modification),Read/Grep(file system access), andget_current_config. - Sanitization: Absent; the skill does not specify any validation, filtering, or escaping of the user-provided input before it is analyzed by the agent.
Audit Metadata