cli-anything-openscreen
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external video recordings and JSON-based project files. This creates a surface for indirect prompt injection if the agent is directed to parse or act upon untrusted metadata or content embedded within these files.\n
- Ingestion points: Reads source video files (e.g.,
recording.mp4) and project files (.openscreen).\n - Boundary markers: None identified in the prompt instructions to the agent.\n
- Capability inventory: Performs video rendering (
export render), frame extraction (media thumbnail), and metadata inspection (media probe) using external tools.\n - Sanitization: The skill description does not specify sanitization or validation logic for the content of the media files.\n- [COMMAND_EXECUTION]: The skill operates by executing command-line tools, specifically the
cli-anything-openscreenPython package, and depends on the system-levelffmpegbinary for video processing. It also references an external utilitycli-hubfor preview inspection.\n- [EXTERNAL_DOWNLOADS]: The skill requires the installation of thecli-anything-openscreenPython package from a public registry. As this is the primary tool for the skill and matches the author's identity, it is considered a legitimate resource.
Audit Metadata