The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute a local Python script (scripts/bilibili.py). This is the intended mechanism for the skill's functionality.
[EXTERNAL_DOWNLOADS]: The rehost_images function fetches images from external URLs provided in article content to upload them to Bilibili's CDN. This behavior is documented and includes significant security controls:
SSRF Protection: The script implements _assert_public_url, which resolves hostnames and verifies they do not point to private, loopback, or reserved IP ranges before initiating a connection.
Redirect Handling: A custom _NoRedirect handler is used during image fetching to prevent attackers from using redirects to bypass hostname checks.
[DATA_EXFILTRATION]: While the skill manages sensitive authentication data via the BILIBILI_COOKIES environment variable, it implements safeguards to prevent accidental exposure:
The script uses urllib.request with add_unredirected_header to ensure cookies are not leaked to external sites if the Bilibili API issues a redirect.
Instructions explicitly warn the agent never to print or echo the cookie content.
[SAFE]: Destructive or public-facing actions (publish and delete-draft) are gated by a dry-run mechanism. They require an explicit --confirm flag, which the script only honors if it is the final argument, ensuring human-in-the-loop validation.

bilibili