export

Pass

Audited by Gen Agent Trust Hub on Aug 26, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the retrieval of project exports from remote downloadUrl endpoints generated by the ChatCut hosted connector during the render process.\n- [COMMAND_EXECUTION]: Instructions direct the agent to perform local file system operations, specifically downloading remote assets to collision-safe local paths and managing render-job tracking through tool invocations.\n- [INDIRECT_PROMPT_INJECTION]: The skill defines a surface for indirect prompt injection by ingesting data from external URLs and project metadata without specified sanitization.\n
  • Ingestion points: External media assets and render metadata retrieved from downloadUrl (SKILL.md).\n
  • Boundary markers: None explicitly defined to delimit or isolate content retrieved from external sources.\n
  • Capability inventory: Writing to the local file system, network communication via submit_export, and media asset management tools (SKILL.md).\n
  • Sanitization: The instructions do not specify validation or filtering requirements for the content or metadata of downloaded assets.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 26, 2026, 03:25 PM
Security Audit — agent-trust-hub — export