news-rough-cut

Pass

Audited by Gen Agent Trust Hub on Nov 5, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data derived from external media, creating a surface for indirect prompt injection where malicious instructions could be embedded in the audio/transcript of the news footage.
  • Ingestion points: The skill uses the transcribe_track tool to ingest and process text from user-provided news footage (referenced in SKILL.md and references/deterministic-execution-and-acceptance.md).
  • Boundary markers: There are no explicit boundary markers or system instructions to ignore potential commands embedded within the transcript data.
  • Capability inventory: The agent has access to a variety of timeline manipulation tools such as edit_item, split_item, edit_track, and read_project, which are used to execute the editing plan based on the ingested content.
  • Sanitization: No content validation or sanitization of the transcript text is implemented before it is used to determine editing points.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 5, 2026, 11:47 AM