agile-router
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided context via the
$ARGUMENTSvariable to determine which agile skill to recommend. This represents a standard ingestion surface for untrusted data, although the impact is limited by the skill's restricted capabilities. \n - Ingestion points: Untrusted user input is captured in the
$ARGUMENTSvariable withinSKILL.md. \n - Boundary markers: The prompt logic does not utilize explicit delimiters or 'ignore' instructions to isolate user-provided context from the router's decision logic. \n
- Capability inventory: The skill's operations are confined to logic and recommendations; it lacks the ability to write to the file system, perform network operations, or execute shell commands. \n
- Sanitization: No explicit sanitization or input validation is performed on the
$ARGUMENTScontent before it is processed.
Audit Metadata