extracting-structure

Pass

Audited by Gen Agent Trust Hub on Aug 22, 2026

Risk Level: SAFECOMMAND_EXECUTIONNO_CODE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to execute the watch-skill command for deterministic data extraction. The commands follow a strict syntax and are used as intended for video analysis tasks.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data such as video transcripts and OCR text extracted from recordings.
  • Ingestion points: Data is read from a persistent video index via the watch-skill command (SKILL.md).
  • Boundary markers: No specific delimiters or safety instructions are defined to separate the extracted video content from agent instructions.
  • Capability inventory: The skill has access to Bash and Read tools (SKILL.md frontmatter).
  • Sanitization: The skill does not describe specific filtering or sanitization of the OCR or transcript data before it is presented to the user or used in issue reports.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 22, 2026, 12:04 AM
Security Audit — agent-trust-hub — extracting-structure