parallel-agent
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates multi-agent orchestration where a parent agent ingests findings and status reports from child agents, creating a data ingestion surface.
- Ingestion points: Data enters the parent context through sub-agent reports including status, findings, and changed files as defined in SKILL.md.
- Boundary markers: The skill prescribes using 'fresh' lanes for isolation and requires the parent agent to synthesise and validate all results rather than adopting them directly.
- Capability inventory: The orchestrated sub-agents have capabilities for file modification and command execution via the pi-subagents dependency.
- Sanitization: Security is maintained through mandatory validation gates and human-in-the-loop synthesis where the parent agent must review all changed files and residual risks.
Audit Metadata