gen-ai-use
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill documentation includes an installation command that downloads a script from the official vendor domain (https://picsart.com/gen-ai-cli/install.sh) and executes it using bash. As this originates from the verified author domain, it is considered standard vendor practice.
- [COMMAND_EXECUTION]: The skill functions by invoking the gen-ai CLI tool through shell commands. It passes user-defined arguments such as prompts, model IDs, and local file paths to the executable.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external media files and manifest configurations, which constitutes a surface for indirect prompt injection.
- Ingestion points: The agent reads local media files (images, videos, audio) and job manifests (manifest.json) to pass as inputs to AI models.
- Boundary markers: The instructions do not specify delimiters or safety warnings to prevent the model from following instructions embedded within these external inputs.
- Capability inventory: The skill uses the Bash tool to execute the CLI, which performs network operations to the Picsart API and interacts with Picsart Drive for file storage.
- Sanitization: There is no description of sanitization or validation logic applied to the content of the input files before processing.
Audit Metadata