Bilibili Downloader

Download videos, audio, subtitles, and covers from Bilibili using bilibili-api. Use when working with Bilibili content for downloading videos in various qual...

This package appears to implement a normal Bilibili downloader, but there are a few small inconsistencies you should check before running or providing secrets: - The SKILL.md suggests exporting BILIBILI_SESSDATA (a browser cookie) for premium content, but the provided scripts do not read that env var or consume cookies from the provided config file. Do not paste your real SESSDATA into config files or environment variables until you confirm where it's used. Prefer running without credentials for public content. - The bundled scripts use the third-party package bilibili-api-python (install via pip). Inspect that library's source or install it in an isolated environment (virtualenv/container) before running. - The package metadata in _meta.json has a different ownerId than the registry metadata shown; this could mean repackaging or inconsistent metadata—verify the source and trustworthiness of the skill before use. - There are no remote installers in the skill, which reduces risk. Still: run the scripts locally in a sandbox (or review them line-by-line) and verify they behave as expected. If you must download private/premium content, only provide authentication tokens/cookies to code you trust; rotate cookies after testing. If you want higher confidence, ask the publisher for the canonical homepage or repo, confirm the bilibili-api-python package name and its maintainers, and request that the skill explicitly document and/or use credentials in code (so you can review handling) rather than only in prose.

Type: OpenClaw Skill Name: sunshine-bilibili-downloader Version: 1.0.1 The skill bundle is a legitimate tool for downloading Bilibili content (videos, audio, and subtitles) using the well-known 'bilibili-api-python' library. The provided Python scripts (e.g., scripts/download_video.py, scripts/download_playlist.py) and documentation (SKILL.md) are well-structured, transparent, and align perfectly with the stated purpose without any signs of malicious intent, data exfiltration, or harmful prompt injection.