repurpose-for-channels
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides the agent with specific command strings to facilitate the setup of the Clueso Model Context Protocol (MCP) server, such as
claude mcp add --transport http Clueso https://connect.clueso.io/mcp. These commands allow the agent to integrate with the vendor's external video editing tools and services. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data in the form of video transcripts and rendered frames to determine narrative arcs and key moments for social cuts.
- Ingestion points: Video transcripts and scene frames from Clueso projects or user-uploaded raw recordings (SKILL.md, Workflow Step 1).
- Boundary markers: None identified; the agent is instructed to watch and analyze the transcript directly to find narrative arcs.
- Capability inventory: Project duplication, video trimming, scene resizing, captioning, and video export operations via the Clueso API.
- Sanitization: No explicit sanitization or filtering of transcript text is mentioned; the agent relies on the transcript content to identify the "strongest narrative."
Audit Metadata