Item: bilibili-cli
Rating: 7.8
Author: Implexa

bilibili-cli

CLI skill for Bilibili (哔哩哔哩, B站) with token-efficient YAML output for AI agents to browse videos, users, search, trending, dynamics, favorites, and interact...

copies: implexa run clawhub/bilibili-cli

view source

installs

stars

karma

SkillRank score ↗

7.8/ 10

evaluated by implexa, claude-haiku-4-5 · 2026-05-26

bilibili-cli provides a comprehensive command-line interface for querying bilibili video metadata, user profiles, search results, and interactions. it outputs yaml and json for ai agents and includes workflows for content analysis and user research.

structure

9.0

trigger phrases

8.0

procedure

8.0

edge cases

7.0

documentation

8.0

strengths

view original SKILL.md from clawhubclick to expand

---
name: bilibili-cli
description: CLI skill for Bilibili (哔哩哔哩, B站) with token-efficient YAML output for AI agents to browse videos, users, search, trending, dynamics, favorites, and interactions from the terminal
author: jackwener
version: "1.0.0"
tags:
  - bilibili
  - 哔哩哔哩
  - b站
  - video
  - social-media
  - cli
---

# bilibili-cli Skill

A CLI tool for interacting with Bilibili (哔哩哔哩). Use it to fetch video info, search content, browse user profiles, and perform interactions like liking or triple-clicking.

## Agent Defaults

When you need machine-readable output:

1. Prefer `--yaml` first because it is usually more token-efficient than pretty JSON.
2. Use `--json` only when downstream tooling strictly requires JSON.
3. Keep result sets small with `--max`, `--page`, or `--offset`.
4. Prefer specific commands over broad ones. Example: use `bili user-videos 946974 --max 3 --yaml` instead of fetching large timelines.
5. When summarizing a video, fetch subtitles first. Subtitles usually contain the video's core content and are the best primary source for summaries.
6. Only fall back to `--ai`, comments, or audio extraction when subtitles are unavailable or clearly insufficient.

## Prerequisites

```bash
# Install (requires Python 3.10+)
uv tool install bilibili-cli
# Or: pipx install bilibili-cli

# If you need audio extraction support (requires PyAV)
uv tool install "bilibili-cli[audio]"
# Or: pipx install "bilibili-cli[audio]"
```

## Authentication

Most read commands work without login. Subtitles, favorites/following/watch-later/history, feed, and interactions require login.

```bash
bili status                    # Check if logged in (exit 0 = yes, 1 = no)
bili login                     # QR code login (if not authenticated)
```

Authentication auto-detects local browser cookies (Chrome/Firefox/Edge/Brave). If cookies are found and valid, no manual login needed. Credentials are saved to `~/.bilibili-cli/credential.json`.

## Command Reference

### Video

```bash
# Get video details (accepts BV ID or full URL)
bili video BV1ABcsztEcY
bili video https://www.bilibili.com/video/BV1ABcsztEcY

# Options
bili video BV1ABcsztEcY --subtitle            # Show subtitles (plain text)
bili video BV1ABcsztEcY --subtitle-timeline   # Show subtitles with timestamps
bili video BV1ABcsztEcY -st --subtitle-format srt  # Export as SRT format
bili video BV1ABcsztEcY --ai            # Show B站 AI summary
bili video BV1ABcsztEcY --comments      # Show top comments
bili video BV1ABcsztEcY --related       # Show related videos
bili video BV1ABcsztEcY --yaml          # Token-efficient YAML output
bili video BV1ABcsztEcY --json          # Structured JSON envelope
```

### User

```bash
# Look up user profile (by UID or username)
bili user 946974
bili user "影视飓风"

# List user's videos
bili user-videos 946974 --max 20
bili user-videos "影视飓风" --yaml
```

### Search

```bash
# Search users (default)
bili search "关键词"

# Search videos
bili search "关键词" --type video

# Pagination and limit
bili search "关键词" --type video --max 5
bili search "关键词" --page 2
```

### Discovery

```bash
bili hot                       # Trending/popular videos
bili hot --page 2 --max 10     # Page 2, limit 10
bili rank                      # Site-wide ranking (3-day)
bili rank --day 7 --max 30     # 7-day ranking, top 30
bili feed                      # Dynamic timeline (requires login)
bili feed --offset 1234567890  # Next page via returned cursor
bili my-dynamics               # My posted dynamics (requires login)
bili dynamic-post "hello"      # Publish text dynamic (requires write credential)
bili dynamic-delete 123456789  # Delete one dynamic (requires write credential)
```

### Collections (require login)

```bash
bili favorites                 # List favorite folders
bili favorites <ID> --page 2   # Videos in a folder
bili following                 # Following list
bili watch-later               # Watch later list
bili history                   # Watch history
```

### Audio Extraction

Requires `bilibili-cli[audio]` extra (PyAV). Install with `uv tool install "bilibili-cli[audio]"`.

```bash
# Download audio and split into ASR-ready WAV segments (25s each, 16kHz mono)
bili audio BV1ABcsztEcY                 # Split to /tmp/bilibili-cli/{title}/
bili audio BV1ABcsztEcY --segment 60    # 60s per segment
bili audio BV1ABcsztEcY --no-split      # Full m4a file, no splitting
bili audio BV1ABcsztEcY -o ~/data/      # Custom output directory
```

### Interactions (require login)

```bash
bili like BV1ABcsztEcY         # Like a video
bili like BV1ABcsztEcY --undo  # Unlike
bili coin BV1ABcsztEcY         # Give 1 coin
bili coin BV1ABcsztEcY -n 2    # Give 2 coins
bili triple BV1ABcsztEcY       # 一键三连 (like + coin + favorite)
bili unfollow 946974           # Unfollow by UID
```

### Account

```bash
bili status                    # Quick login check
bili status --yaml             # Structured auth status
bili whoami                    # Detailed profile info
bili whoami --yaml              # Profile as YAML
bili whoami --json              # Profile as JSON
bili login                     # QR code login
bili logout                    # Clear credentials
```

## Structured Output

Major query commands support both `--yaml` and `--json` for machine-readable output. Prefer YAML for agent use:

```bash
bili status --yaml                                # Quick structured auth check
bili video BV1ABcsztEcY --yaml                       # Preferred for AI agents
bili hot --max 5 --yaml                             # Smaller, token-efficient payload
bili user 946974 --json | jq -r '.data.user.name'   # JSON when jq is needed
```

When stdout is not a TTY, `bilibili-cli` defaults to YAML automatically.
Use `OUTPUT=yaml|json|rich|auto` to override the default output mode.
All machine-readable output uses the envelope documented in [SCHEMA.md](./SCHEMA.md).

## Debugging

```bash
bili -v <command>              # Enable verbose/debug logging for any command
```

## Common Patterns for AI Agents

```bash
# For video summarization, fetch subtitles first
bili video BV1ABcsztEcY --subtitle

# Only use AI summary as a fallback or secondary signal
bili video BV1ABcsztEcY --ai

# Get comments for sentiment analysis
bili video BV1ABcsztEcY --comments

# Extract audio for speech-to-text (ASR)
# Segments are saved to /tmp/bilibili-cli/{title}/seg_000.wav, seg_001.wav, ...
bili audio BV1ABcsztEcY --segment 25

# Find a user's latest video BV ID with minimal payload
bili user-videos 946974 --max 1 --yaml

# Check if logged in before performing actions
bili status && bili like BV1ABcsztEcY

# Search and inspect the first few results
bili search "topic" --type video --max 3 --yaml
```

### Workflow: Video Content Analysis

```bash
# 1. Search for a topic
bili search "AI" --type video --max 5

# 2. Get subtitles first for summarization
bili video BV1xxx --subtitle

# 3. If subtitles are missing or incomplete, try AI summary
bili video BV1xxx --ai

# 4. If there is still not enough content, extract audio for ASR
bili audio BV1xxx --segment 25

# 5. Get comments for audience reaction
bili video BV1xxx --comments
```

### Workflow: UP主 Research

```bash
# 1. Look up UP主 profile
bili user "影视飓风"

# 2. Get their recent videos
bili user-videos 946974 --max 10

# 3. Inspect a specific video
bili video BV1xxx --ai --comments
```

## Error Handling

- Commands exit with code 0 on success, non-zero on failure
- Error messages are prefixed with ❌
- Login-required commands show ⚠️ with instruction to run `bili login`
- Invalid BV IDs show a clear error message

## Safety Notes

- Do not ask users to share raw credential/cookie values in chat logs.
- Prefer local browser cookie extraction over manual secret copy/paste.
- If auth fails, ask the user to re-login via `bili login`.

don't have the plugin yet? install it then click "run inline in claude" again.

added explicit inputs with auth details and external connections, structured procedure into 15 numbered steps with clear input/output, extracted 10 decision points from implicit logic throughout original doc, documented output contract with YAML/JSON envelopes and file locations, specified outcome signals with exit codes and confirmation patterns, and added edge cases for rate limits auth expiry and empty result sets.

intent

use bilibili-cli to query Bilibili (哔哩哔哩) for video metadata, user profiles, search results, trending content, and perform authenticated interactions (likes, coins, follows) from the terminal. most read operations work unauthenticated, but subtitles, favorites, history, feed, and interactions require login. prefer YAML output for token efficiency in agentic workflows.

inputs

external connection: bilibili.com

unauthenticated reads: video info, user profiles, search, trends, rankings
authenticated reads: subtitles, favorites, watch-later, history, feed, dynamics
write operations: like, coin, follow, post dynamic

authentication (auto-detected or manual)

local browser cookies from Chrome, Firefox, Edge, or Brave (auto-extracted on first run)
fallback: QR code login via bili login
credentials stored in ~/.bilibili-cli/credential.json
check status with bili status (exit code 0 = logged in, 1 = not)

dependencies

Python 3.10+
bilibili-cli package (install via uv tool install bilibili-cli or pipx install bilibili-cli)
optional: PyAV for audio extraction (uv tool install "bilibili-cli[audio]")

environment variables (optional)

OUTPUT: override default output format (yaml, json, rich, auto). defaults to yaml when stdout is not a TTY

procedure

install bilibili-cli
- input: python 3.10+, uv or pipx
- run: uv tool install bilibili-cli (or pipx install bilibili-cli)
- output: bilibili-cli command available in shell
check authentication status
- input: none
- run: bili status or bili status --yaml
- output: exit code 0 (logged in) or 1 (not logged in); YAML/text status
- decision: if exit code 1, proceed to step 3; else skip to step 4
authenticate (if needed)
- input: none
- run: bili login
- output: QR code displayed in terminal (scan with phone). credentials saved to ~/.bilibili-cli/credential.json
- fallback: if QR code auth fails, browser cookies auto-detected on next command
- outcome: bili status returns exit code 0
query videos (by BV ID or URL)
- input: BV ID (e.g. BV1ABcsztEcY) or full URL
- run: bili video BV1ABcsztEcY --yaml (add --subtitle, --ai, --comments, --related as needed)
- output: YAML-formatted video metadata (title, views, likes, description, publish date)
- note: subtitles, AI summary, comments require login; plain video info does not
fetch subtitles (preferred for summarization)
- input: BV ID
- run: bili video BV1ABcsztEcY --subtitle (or --subtitle-timeline for timestamps, --subtitle-format srt for SRT export)
- output: plain text or SRT format subtitles
- decision: if subtitles exist, use them; else fallback to step 6
- note: requires login
fetch AI summary (fallback)
- input: BV ID
- run: bili video BV1ABcsztEcY --ai
- output: Bilibili AI-generated summary text
- note: only if subtitles unavailable or insufficient; requires login
extract audio (if subtitle/AI insufficient)
- input: BV ID, optional segment duration (default 25s), optional output directory
- run: bili audio BV1ABcsztEcY --segment 25 -o ~/data/
- output: WAV files (16kHz mono, 25s segments) in output directory or /tmp/bilibili-cli/{title}/
- note: requires bilibili-cli[audio] extra; files named seg_000.wav, seg_001.wav, etc.
search videos or users
- input: keyword, optional type (default=user; video, user, etc.), optional max results, optional page
- run: bili search "keyword" --type video --max 5 --yaml
- output: YAML list of matching videos or users
- note: no login required
browse user profile and videos
- input: UID (numeric) or username (string)
- run: bili user 946974 (or bili user "username")
- run for videos: bili user-videos 946974 --max 20 --yaml
- output: YAML user profile data and video list
- note: no login required for public profiles
explore trending/ranking content
- input: optional page, optional max results, optional day range (for rank)
- run: bili hot --max 10 --yaml (or bili rank --day 7 --max 30)
- output: YAML list of trending/ranked videos
- note: no login required
fetch feed (authenticated users)
- input: optional offset cursor from previous result
- run: bili feed or bili feed --offset 1234567890
- output: YAML timeline of followed creators' videos and updates
- note: requires login
access collections (authenticated users)
- input: collection type (favorites, watch-later, history); optional folder ID or pagination
- run: bili favorites (or bili favorites <ID> --page 2), bili watch-later, bili history
- output: YAML list of videos in collection
- note: requires login
perform interaction (like, coin, follow)
- input: BV ID (for video interactions) or UID (for follow/unfollow)
- run: bili like BV1ABcsztEcY (or --undo to unlike), bili coin BV1ABcsztEcY -n 2, bili triple BV1ABcsztEcY, bili unfollow 946974
- output: confirmation message or YAML success envelope
- decision: if already liked/followed, check for --undo flag
- note: requires login and write credential
post or delete dynamic (authenticated)
- input: text content (for post) or dynamic ID (for delete)
- run: bili dynamic-post "hello" or bili dynamic-delete 123456789
- output: confirmation or YAML envelope
- note: requires login and write credential
debug/troubleshoot
- input: any command + -v flag
- run: bili -v video BV1ABcsztEcY
- output: verbose debug logs to stderr
- note: use for diagnosing API errors, network timeouts, auth failures

decision points

if not logged in (exit code 1 from step 2)

then: run bili login to authenticate via QR code or auto-detect browser cookies
else: proceed to querying/interacting

if command requires login but user is not authenticated

then: CLI shows warning ⚠️ with instruction to run bili login
else: command executes normally

if subtitles are available for a video

then: fetch and use subtitles as primary source for summarization (most token-efficient, most accurate)
else: fallback to --ai summary if available
else: fallback to --comments or audio extraction

if API rate-limit hit (429 response)

then: retry with exponential backoff (recommend wait 60+ seconds before retry)
else: proceed normally
note: Bilibili limits requests per IP and per user session; batch operations should respect delays

if authentication token expired or invalid

then: CLI shows ⚠️ prompt, user must re-run bili login to refresh
else: command executes with current credential

if YAML output is not a TTY (piped, redirected, or non-interactive)

then: bilibili-cli defaults to YAML format automatically
else (interactive terminal): defaults to rich/pretty output unless --yaml or --json specified

if BV ID is invalid or video does not exist

then: CLI returns error message with 404 or "video not found"
else: proceed normally

if result set is empty (no videos found, no users matching search, empty feed)

then: CLI returns empty YAML array or null, exits with code 0 (not an error)
else: proceed with results

if network timeout or connection refused

then: CLI shows error ❌, exits with non-zero code
retry by re-running command; consider checking internet connectivity
else: command completes normally

output contract

structured output formats (YAML and JSON)

all major query commands support --yaml and --json flags. output follows a standard envelope (documented in SCHEMA.md).

YAML example (preferred for agents):

code: 0
message: success
data:
  video:
    bvid: BV1ABcsztEcY
    title: "Video Title"
    views: 12345
    likes: 678
    publish: "2024-01-15"
    description: "..."
    subtitles:
      - text: "subtitle line 1"
        timestamp: "00:00:05"
      - text: "subtitle line 2"
        timestamp: "00:00:10"

JSON example:

{
  "code": 0,
  "message": "success",
  "data": {
    "video": {
      "bvid": "BV1ABcsztEcY",
      "title": "Video Title",
      "views": 12345
    }
  }
}

file locations

credentials: ~/.bilibili-cli/credential.json
audio segments: /tmp/bilibili-cli/{video_title}/seg_000.wav, seg_001.wav, etc. (or custom -o directory)
logs: stderr (when -v flag used)

exit codes

0: success
1: not logged in (when login required)
non-zero: command failed (invalid BV ID, API error, network error, etc.)

outcome signal

successful query commands print machine-readable output

bili video BV1ABcsztEcY --yaml outputs YAML envelope with code: 0 and video metadata
bili search "keyword" --type video --max 5 --yaml outputs list of videos
bili user 946974 --yaml outputs user profile
exit code is 0

successful authentication

bili login displays QR code, user scans on phone, credential saved
bili status returns exit code 0 and prints "✓ logged in" (or similar)
bili whoami --yaml shows authenticated user profile

successful interaction (like, coin, follow, post)

CLI prints confirmation (e.g., "✓ video liked" or YAML envelope with code: 0)
exit code is 0

successful audio extraction

WAV files appear in output directory /tmp/bilibili-cli/{title}/ with names seg_000.wav, seg_001.wav, etc.
file sizes match expected segment durations (25s at 16kHz mono = ~800KB per segment, etc.)
exit code is 0

error conditions show clear feedback

invalid BV ID: ❌ error message printed, exit code non-zero
login required: ⚠️ warning with instruction to run bili login, exit code 1
network timeout: ❌ error message, exit code non-zero (retry advised)
rate limit: ❌ message (suggest waiting before retry)
empty result: YAML empty array or null, exit code 0 (not an error)

agent integration patterns

parse YAML with yaml library (Python) or equivalent
check code field in response: 0 = success, non-zero = error
extract data payload for downstream processing
respect exit codes in shell pipelines (e.g., bili status && bili like BV1...)

bilibili-cli

related skills

intent

inputs

procedure

decision points

output contract

outcome signal