ai302-cli

Pass

Audited by Gen Agent Trust Hub on Jun 4, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install the cli_302ai Python package. This is a vendor-owned resource originating from the author '302ai' and is used for its intended media generation functionality.
  • [COMMAND_EXECUTION]: The skill relies on executing various subcommands of the 302ai CLI tool (e.g., image, video, tts, search). These commands facilitate interaction with external AI models and services.
  • [PROMPT_INJECTION]: The skill exhibits a potential surface for indirect prompt injection when processing external content.
  • Ingestion points: The skill ingests untrusted data from the web via the search module (search.md) and from transcribed audio files via the stt module (stt.md).
  • Boundary markers: There are no explicit instructions to use delimiters or "ignore previous instructions" markers when handling results from these external sources.
  • Capability inventory: The skill possesses extensive command execution capabilities across all its modules (image, video, 3D, song, search, etc.) which could be misused if the agent obeys instructions hidden in search results.
  • Sanitization: No specific sanitization, escaping, or validation logic is defined for processing the outputs of search queries or transcription tasks before they are returned to the agent's context.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 4, 2026, 11:54 AM
Security Audit — agent-trust-hub — ai302-cli