video-memory
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
Bashtool to execute a local CLI utility,watch-skill. This tool is used to list indexed videos, perform keyword/semantic searches, and synthesize answers from video data stored at~/.watch-skill/. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from untrusted sources (video frames, OCR, and transcripts).
- Ingestion points: Data enters the agent context through the outputs of
watch-skill search,watch-skill ask, andwatch-skill library askcommands executed inSKILL.md. - Boundary markers: Absent. The instructions do not provide delimiters or "ignore embedded instructions" warnings for the agent when interpreting search results or video transcripts.
- Capability inventory: The skill has access to the
Bashtool, allowing it to execute system commands based on instructions derived from the indexed content. - Sanitization: Absent. There is no mention of validation, filtering, or escaping of the content retrieved from the persistent video index before it is used by the agent.
Audit Metadata