meeting-insights-analyzer

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect prompt injection surface detected via meeting transcript analysis.
  • Ingestion points: The skill scans directories for .txt, .md, .vtt, .srt, and .docx files as outlined in the instructions.
  • Boundary markers: The skill lacks explicit instructions to treat transcript content as untrusted or to use delimiters to prevent instruction following from within the data.
  • Capability inventory: The agent reads file content, performs behavioral analysis, and generates output containing direct quotes from the transcripts.
  • Sanitization: No input validation or sanitization is applied to the ingested transcript text before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 06:46 PM