skill-bilibili-upload · Bilibili Video Upload

A Bilibili upload skill wrapping the mature biliup CLI: scan a QR code once to log in and generate a cookie, then reuse it. It supports title, description, category, tags, cover image, repost declarations, and scheduled publishing — defaulting to a dry-run preview of the upload command before anything really goes out.

Every successful upload is followed by automatic read-back verification (matching new submissions by title prefix + time window); only verified uploads count as success, and unverified ones are reported clearly with a prompt for manual confirmation.

Example invocation: "Upload out.mp4 to Bilibili's knowledge category."

Full brief

Positioning

skill-bilibili-upload is a Bilibili submission channel wrapping the mature biliup CLI: log in once via QR code, reuse the cookie afterwards. It exposes the full set of upload parameters (category, tags, cover, repost declaration, scheduling), with dry-run previews and read-back verification guarding publish quality. It handles uploading only — no video editing.

Core capabilities

• QR login + cookie reuse: first login generates cookies.json via QR scan; reused afterwards.
• Full upload parameters: title, description, category (30 built-in Chinese category names mapped, or --tid number), tags (≤ 10), cover, repost declaration (--copyright 2 --source), scheduled publishing (--dtime 10-digit timestamp, must be > 4 hours out).
• Dry-run preview: omitting --exec previews the upload command and parameters; add --exec to actually submit.
• Pre-submit sensitive-data scan: title, description, tags, and repost source are force-scanned before a real upload; secrets, internal addresses, or file paths block submission.
• Read-back verification: after upload, the skill queries the submission API directly to find the new video by title prefix + time window, with four outcome grades; unverified results are reported clearly with a manual-check prompt — never counted as success.
• Publish logging: on success it calls skill-publish-log to record platform, time, title, and content ID; failures are never logged as success.
• Profile-aware: with a profile, default category/tags follow the profile's Bilibili positioning and title style matches the persona.

Workflow

1. check confirms biliup and cookie readiness
2. First use: login QR scan; tid lists categories
3. upload (no --exec) previews the upload command and parameters
4. Add --exec to submit for real
5. Automatic read-back verification of the new submission
6. On success, record via skill-publish-log

Inputs & outputs

Output: upload result + read-back verdict (success, or unverified with manual-check guidance).

Boundaries with adjacent skills

• skill-kuaishou-upload: uses Playwright browser automation, focused on vertical 9:16; this skill uses the biliup CLI, with horizontal 16:9 as the Bilibili norm.
• This skill supports scheduled publishing and repost declarations, plus a pre-submit sensitive-data scan.

Fit

• Distributing knowledge/tutorial long-form video on Bilibili
• Submissions that need scheduled timing
• Reposts requiring a repost declaration
• Previewing parameters before batch submissions

Before you start

• Install the biliup CLI (pip install biliup); login and submission need internet access to Bilibili.
• First use requires QR login to generate the cookie; cookies are sensitive — store them safely, never leak.
• Bilibili skews horizontal 16:9; convert vertical footage first with video-reframe.
• A wrong category hurts recommendations — confirm with the user when unsure; cover and tags (≤ 10) materially affect clicks and recommendations, so fill them in.

skill-bilibili-upload is part of the Aiglade Skill library. Invoke it from the Aiglade chat box in plain language.

← All 155 skills