feature-planning

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • PROMPT_INJECTION (LOW): The skill contains an indirect prompt injection surface (Category 8) because it reads external project files and interpolates their content into prompts for downstream skills without boundary markers or sanitization.\n
  • Ingestion points: Reads CLAUDE.md (Step 1) and all Z01 research files (Step 3) found in the repository.\n
  • Boundary markers: Absent. The content extracted from these files is directly embedded into the prompt for the superpowers:writing-plans skill using simple string interpolation.\n
  • Capability inventory: The skill has permissions to read arbitrary project files, invoke the superpowers:writing-plans skill, and write implementation plans and clarification files to the filesystem.\n
  • Sanitization: No sanitization, escaping, or validation of the ingested file content is performed before passing it to the planning agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:06 PM