parallel-delivery

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill consists entirely of documentation and procedural instructions; it does not include any scripts, commands, or executable code.\n- [SAFE]: The described workflow enforces security best practices, including separation of duties (mandatory review by a different session or person) and workspace isolation (using fresh clones or worktrees) to prevent data collisions.\n- [SAFE]: The instructions reduce risks from under-specified tasks by requiring agents to halt and report gaps rather than guessing, which helps prevent accidental implementation of insecure features based on assumptions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 02:20 AM
Security Audit — agent-trust-hub — parallel-delivery