parallel-execution

Pass

Audited by Gen Agent Trust Hub on Feb 24, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill is purely educational, offering structural templates for improving task throughput without performing any programmatic operations.
  • [NO_CODE]: The skill consists exclusively of documentation and examples in markdown; no executable scripts, shell commands, or binary files are included.
  • [SAFE]: No external network requests, data exfiltration patterns, or hardcoded credentials were found within the instructions.
  • [SAFE]: Analysis of indirect prompt injection surface (Category 8): 1. Ingestion points: The prompt parameter in the runSubagent tool templates in SKILL.md. 2. Boundary markers: Not utilized in the instructional examples. 3. Capability inventory: Orchestrates subagent execution via runSubagent. 4. Sanitization: Not applicable as the skill does not implement data processing logic, focusing instead on execution flow optimization.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 24, 2026, 11:09 PM