scenario-war-room
Pass
Audited by Gen Agent Trust Hub on Apr 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the user or agent to execute a local Python script located at
scripts/scenario_modeler.pyfor interactive modeling. - [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection within its Multi-Agent Analysis framework. 1. Ingestion points: Untrusted data enters the context via variables such as
{risk_a},{risk_b}, and{company}within theTaskprompts in SKILL.md. 2. Boundary markers: Absent; these variables are interpolated directly into the sub-agent instructions without delimiters or clear boundary definitions. 3. Capability inventory: The skill has the ability to spawn sub-agents and execute local shell commands/scripts. 4. Sanitization: Absent; there is no visible mechanism to escape or validate external content before it is processed by sub-agents.
Audit Metadata