chaptered-training-video

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references several URLs under the clueso.io domain (e.g., https://connect.clueso.io/mcp, https://help.clueso.io/mcp-setup). These are official resources used for service authentication and documentation.
  • [COMMAND_EXECUTION]: The instructions include a command for the assistant to set up the environment: claude mcp add --transport http Clueso https://connect.clueso.io/mcp. This is a standard configuration step for the Model Context Protocol (MCP) used by this vendor.
  • [PROMPT_INJECTION]: The skill analyzes external data (video transcripts) to identify chapter boundaries. While this represents a potential surface for indirect prompt injection if a transcript contained malicious instructions, the skill's capabilities are focused on video editing operations (splitting, naming, and adding title cards), which limits the impact of such an attack.
  • Ingestion points: Audio transcription and video metadata analysis (Workflow Step 2).
  • Boundary markers: None explicitly defined for transcript analysis.
  • Capability inventory: Project manipulation, clip creation, and timeline splitting via Clueso tools.
  • Sanitization: Not explicitly documented.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 09:42 AM
Security Audit — agent-trust-hub — chaptered-training-video