The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Retrieves model pricing, usage data, and request metadata from the official fal.ai API (api.fal.ai). These requests are necessary for the skill's functionality and target a well-known technology service.
[COMMAND_EXECUTION]: Employs curl to perform API operations and python3 to parse the resulting JSON data.
User-provided inputs, such as model identifiers, are sanitized using URL encoding.
Data is passed to Python logic using positional arguments, ensuring that external content is treated as data rather than executable code.
[CREDENTIALS_UNSAFE]: Includes a setup utility (setup.sh) for managing the FAL_KEY credential. The key is stored in a local .env file, which is a standard pattern for local tools. The script includes basic validation of the key's format and does not contain hardcoded secrets.
[SAFE]: No malicious patterns, such as prompt injection, obfuscation, or unauthorized data exfiltration, were detected. The skill's behavior is consistent with its stated purpose of assisting with fal.ai platform management.

fal-platform