omnicaptions-download

Pass

Audited by Gen Agent Trust Hub on Sep 22, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructions specify the installation of the omni-captions-skills package from a vendor-managed repository at https://lattifai.github.io/pypi/simple/. This is a resource owned by the skill author.
  • [COMMAND_EXECUTION]: The skill uses a specialized CLI tool, omnicaptions, to perform downloads and media processing via Bash commands. The scope is limited to the omnicaptions toolset.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted URLs provided by users or found in data, which could potentially contain malicious payloads.
  • Ingestion points: External URLs and platform-specific video IDs are passed as arguments to the download command.
  • Boundary markers: The skill explicitly requires the agent to display the URL, quality, and path, and obtain user confirmation before proceeding.
  • Capability inventory: The skill can execute shell commands within the omnicaptions namespace and write files (media, captions, metadata) to the filesystem.
  • Sanitization: Validation is performed by the underlying yt-dlp library, and the agent is instructed not to proceed without human approval.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 22, 2026, 03:49 AM