ticket-topics-to-video-batch
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires connecting to an external Model Context Protocol (MCP) endpoint at
https://connect.clueso.io/mcp. This is a functional requirement for the agent to access the vendor's video creation tools. - [COMMAND_EXECUTION]: The instructions include platform-specific setup commands, such as
claude mcp add --transport http Clueso https://connect.clueso.io/mcp, to enable the necessary tool integrations within the assistant environment. - [DATA_EXFILTRATION]: The skill handles potentially sensitive support ticket data, including resolution recordings and descriptions. It mitigates exposure risks by explicitly instructing the agent to perform PII redaction and generalize customer-specific details before outputting content.
- [PROMPT_INJECTION]: The skill processes untrusted input from support tickets, creating a surface for indirect prompt injection (Category 8). \n
- Ingestion points: Support ticket topic lists and written descriptions (SKILL.md). \n
- Boundary markers: The workflow does not specify explicit markers to isolate untrusted ticket content from the instruction set. \n
- Capability inventory: Workspace management and video asset generation via the Clueso MCP tools. \n
- Sanitization: While PII redaction is mandated for privacy, no technical sanitization is described to prevent the execution of instructions embedded within the ticket source material.
Audit Metadata