The Agent Skills Directory

[SAFE]: The skill uses established best practices for handling API credentials, loading them from environment variables or .env files and providing explicit instructions to the agent not to print, inspect, or commit secret values.
[EXTERNAL_DOWNLOADS]: The scripts interact exclusively with official Google YouTube Data API endpoints (https://www.googleapis.com/youtube/v3). These are well-known and trusted services.
[COMMAND_EXECUTION]: Command examples provided in the documentation are for running the included Python scripts using standard tools like uv or python3. The Python client implementation explicitly restricts operations to HTTP GET requests, ensuring the tool remains read-only.
[PROMPT_INJECTION]: Instructions are focused on operational parameters, research methodology, and data reporting; no attempts to bypass safety filters or override agent behavior were found.
[DATA_EXFILTRATION]: Network operations are limited to authenticated requests to the YouTube API. No patterns of sensitive local file access (such as SSH or AWS credentials) or unauthorized data transmission were detected.
[SAFE]: No obfuscation techniques, such as Base64-encoded commands, zero-width characters, or homoglyphs, are present in the code or instructions.

youtube-research