The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill configuration downloads and executes the fal-ai-mcp-server package using npx. This is an expected component for the functionality provided and originates from the official registry for the service.
[COMMAND_EXECUTION]: The skill utilizes npx to launch the MCP server as part of its setup instructions.
[DATA_EXFILTRATION]: The skill includes code snippets that send data to api.elevenlabs.io. This is a legitimate use of a well-known service for text-to-speech synthesis and does not involve the transfer of sensitive local files.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface due to the following factors:
Ingestion points: The skill processes content from external media URLs (image_url, video_url in SKILL.md) and local files via the upload tool.
Boundary markers: The instructions lack explicit delimiters or warnings to the model to ignore potential instructions embedded within the processed media or its metadata.
Capability inventory: The skill has access to network operations (requests.post), file uploading, and generative tools.
Sanitization: There is no mention of content validation or sanitization for data retrieved from external sources before processing.

fal-ai-media