The Agent Skills Directory

[SAFE]: The skill documents secure authentication practices by explicitly warning against embedding FAL_KEY in browser bundles and recommending server-side proxies.
[SAFE]: The skill provides robust security guidance for webhooks, including full code examples for ED25519 signature verification against the official fal.ai JWKS endpoint (https://rest.fal.ai/.well-known/jwks.json).
[EXTERNAL_DOWNLOADS]: The skill references official and well-known packages including @fal-ai/client (NPM) and fal-client (PyPI). These are standard libraries for the service described.
[COMMAND_EXECUTION]: The provided script scripts/verify.sh is a utility for static analysis of the skill's own documentation to ensure compliance with best practices (e.g., checking for deprecated libraries). It does not perform network operations or access sensitive system files.
[CREDENTIALS_UNSAFE]: The skill references FAL_KEY as an environment variable and uses placeholder patterns like key_id:key_secret for documentation. No hardcoded credentials or secrets were found.
[PROMPT_INJECTION]: The skill contains standard instructional content for AI agents. No bypass, override, or system prompt extraction patterns were detected.

fal