parallel-agents

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely documentation-based, providing templates and strategies for agent orchestration without any associated scripts or executable components.
  • [SAFE]: The allowed tools (Read, Glob, Grep) are limited to read-only operations, preventing the skill from performing modifications or data exfiltration.
  • [PROMPT_INJECTION]: While the skill defines roles like 'penetration-tester' and 'security-auditor', it does not contain instructions to bypass AI safety filters or override system constraints. It describes legitimate security review workflows.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 06:59 AM