spawn
Pass
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script
session_manager.pyusing the system shell to update the session state during the spawn process. - [INDIRECT_PROMPT_INJECTION]: The skill builds prompts for subagents by interpolating data from external configuration and template files, creating a vulnerability surface.
- Ingestion points: The skill reads task details and variables from
.agenthub/sessions/{session-id}/config.yamland prompt templates fromreferences/agent-templates.md. - Boundary markers: There are no explicit boundary markers or instructions to ignore embedded commands mentioned in the prompt construction logic.
- Capability inventory: The system can spawn subagents with full filesystem access via
worktreeisolation and execute local Python scripts. - Sanitization: The skill does not describe any validation or sanitization of the content loaded from configuration or template files before it is passed to subagents.
Audit Metadata