product-manager-toolkit

Warn

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: MEDIUMPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION] (MEDIUM): The skill is vulnerable to Indirect Prompt Injection (Category 8) due to its core workflow of processing external, untrusted data.
  • Ingestion points: The customer_interview_analyzer.py script reads interview.txt (transcripts), and rice_prioritizer.py reads features.csv (user/customer feedback).
  • Boundary markers: None. There are no instructions or delimiters defined to prevent the agent from obeying malicious instructions embedded within interview transcripts or feature descriptions.
  • Capability inventory: The skill executes local Python scripts and generates outputs that influence critical business decisions, roadmap generation, and PRD creation.
  • Sanitization: None observed. The markdown does not specify any sanitization or validation of the input text before analysis.
  • [COMMAND_EXECUTION] (LOW): The skill's primary interface involves executing local Python scripts using python3. While this is standard for agent skills, users must ensure the provided scripts (rice_prioritizer.py, etc.) are from a trusted source, as they are executed with the user's local permissions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 16, 2026, 12:40 PM