The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill utilizes the Python requests library to fetch data from the vendor's official API at https://api.scavio.dev. This communication is limited to the intended service and follows standard API integration patterns.
[DATA_EXFILTRATION]: The skill requires the SCAVIO_API_KEY environment variable for authentication. The documentation follows security best practices by advising the user to store this key in environment variables or a secret store rather than hardcoding it in source control.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of external social media data, which is an inherent attack surface for indirect prompt injection.
Ingestion points: Data is ingested through parameters such as user_id, photo_id, url, keyword, and tag defined in SKILL.md.
Boundary markers: The skill provides instructions for handling specific API error codes (e.g., 422, 502) to validate input and handle malformed responses.
Capability inventory: The skill is limited to performing authenticated POST requests to retrieve structured JSON data from the Scavio API.
Sanitization: The instructions advise the agent to maintain original text alongside translations to ensure IDs and handles remain verifiable.
[DYNAMIC_EXECUTION]: The skill documentation contains Python code templates for constructing API requests. These templates involve standard string interpolation for parameters and do not execute untrusted or remote code within the agent's runtime environment.

kuaishou-scraper-api