fal-ai-media
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill interpolates user-provided text directly into prompts for external AI models, creating a vulnerability to indirect prompt injection where malicious instructions could be embedded in user data. * Ingestion points: User-defined prompts for image, video, and audio generation in SKILL.md. * Boundary markers: There are no explicit delimiters or instructions provided to isolate or disregard potential commands within user input. * Capability inventory: The skill utilizes the generate MCP tool and Python requests to send data to external APIs. * Sanitization: No sanitization or filtering logic is described for user input before processing.
- [EXTERNAL_DOWNLOADS]: The skill instructions require fetching the fal-ai-mcp-server package from the NPM registry using the npx utility.
- [COMMAND_EXECUTION]: The skill executes external code via npx as part of the MCP server configuration.
Audit Metadata