writing-plans
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data in the form of project specifications and requirements to generate implementation plans.
- Ingestion points: The SKILL.md file processes external specs to derive tasks, and plan-document-reviewer-prompt.md reviews these documents.
- Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard instructions that might be embedded within the input specifications.
- Capability inventory: The skill generates plans that include shell commands (git, pytest) and writes these plans to the local file system (docs/superpowers/plans/).
- Sanitization: No evidence of sanitization or validation of the input spec content before it is used to generate tasks or code blocks.
- [COMMAND_EXECUTION]: The skill instructions specifically direct the agent to generate and execute shell commands for testing (pytest) and version control (git add, git commit). While these are standard development practices, they constitute an execution capability that could be targeted by malicious instructions in a processed specification.
Audit Metadata