plan-parallel

Pass

Audited by Gen Agent Trust Hub on Mar 7, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill uses strong directives (e.g., 'DO NOT use the EnterPlanMode tool', 'MUST READ') to enforce a specific planning workflow. These are legitimate operational constraints to prevent the agent from skipping steps or entering implementation mode prematurely.
  • [COMMAND_EXECUTION]: The skill utilizes environment-specific slash commands such as /scout, /plan-review, and /plan-validate. These are used for searching the local codebase and validating generated plans, which are standard functions for this type of agent skill.
  • [DATA_EXFILTRATION]: The skill reads project documentation and codebase files to inform its planning. There are no network operations or external data transmissions detected.
  • [REMOTE_CODE_EXECUTION]: The skill is 'planning-only' and explicitly forbids the implementation or execution of code changes. No remote code patterns were found.
  • [NO_CODE]: This is a prompt-based skill provided in markdown format. It does not contain or download executable scripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 7, 2026, 07:59 PM