review-product-plan

Pass

Audited by Gen Agent Trust Hub on Oct 27, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external artifacts such as product plans and requirement documents, which serves as an ingestion point for untrusted data.
  • Ingestion points: The product artifact or linked evidence provided by the user for review.
  • Boundary markers: The skill does not explicitly define delimiters to isolate external content or instruct the agent to ignore instructions embedded within the reviewed data.
  • Capability inventory: The skill uses searching (via MCP) and reading capabilities. It does not perform file-writes, network-exfiltration, or command execution.
  • Sanitization: No explicit sanitization or filtering of input data is mentioned in the instructions.
  • [EXTERNAL_DOWNLOADS]: The 'Provenance' section contains links to external GitHub repositories (e.g., github.com/mekenthompson/ProductOS, github.com/phuryn/pm-skills). These are informational references for the user and do not constitute automated downloads or runtime code execution.
  • [SAFE]: The skill instructions are analytical in nature, guiding the agent to evaluate documents for clarity, grounding, and risk. No evidence of obfuscation, credential theft, or persistence mechanisms was found.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 27, 2026, 03:14 AM