talk-stage5-script

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by reading user-controlled markdown files and interpolating their content into a template for another AI.
  • Ingestion points: Content is read from multiple files in the talks/ directory, including summary, concepts, and angles files.
  • Boundary markers: The skill does not use specific delimiters or instructions to prevent the AI from following instructions that might be embedded within the input talk files.
  • Capability inventory: The skill utilizes Read and Write tools to process and save talk-related files.
  • Sanitization: There is no evidence of content validation or sanitization to ensure that input data does not contain malicious instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 01:07 AM