writing-plans
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to create implementation plans that include meta-instructions for future agent sessions. Specifically, the "Plan Document Header" template includes a directive:
> **For Claude:** REQUIRED SUB-SKILL: Use superpowers:executing-plans to implement this plan task-by-task.This establishes a mechanism where data generated by the skill (the plan) contains instructions that influence subsequent agent behavior. - Ingestion points: Implementation plans are saved to
docs/plans/and intended to be read by agents using theexecuting-plansorsubagent-driven-developmentskills. - Boundary markers: The templates do not define explicit delimiters or instructions to treat embedded content as untrusted data.
- Capability inventory: The generated plans involve shell execution for testing (e.g.,
pytest) and version control (e.g.,git). - Sanitization: There are no provided mechanisms for sanitizing the plan content prior to execution in future steps.
Audit Metadata