edit-video
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests external data that could contain malicious instructions.
- Ingestion points: Raw video files (transcripts and metadata) and DaVinci Resolve timeline states.
- Boundary markers: Absent. There are no explicit delimiters to separate user data from instructions.
- Capability inventory: Execution of local scripts (
doctor,rough-cut,resolve-scripting), management of subagents, and writing to project files (CLAUDE.md, manifest). - Sanitization: Absent. No validation or filtering of transcribed text is performed.
- [COMMAND_EXECUTION]: The skill executes several local automation scripts to interact with the operating system and video editing software.
Audit Metadata