Multi-agent search CLI in Python.
通过 CLI 工具帮你在小红书搜索问题、总结B站视频。
uv run -m deepfind.cli "帮我总结https://www.bilibili.com/video/BV1tew5zVEDf 关于话题世界模型、逃出硅谷 、反OpenAI、AMI Labs、两次拒绝Ilya、杨立昆、李飞飞和42的内容" --num-agent 1 --quiet
uv run -m deepfind.cli "上海哪里赏樱人少?搜索小红书" --num-agent 2
python3 -m pip install -e .Install optional ASR dependencies only when you need Bilibili transcription:
python3 -m pip install -e ".[media]"Pre-download the ASR model on Windows (PowerShell) to avoid first-run delay:
hf download Qwen/Qwen3-ASR-1.7B --repo-type modeluv tool install bilibili-cli
uv tool install xiaohongshu-cli
uv tool install twitter-cli
bili whoami
xhs whoami
twitter whoamiThe CLI auto-loads .env from the repo root.
cp .env.example .envMinimal .env:
QWEN_API_KEY=...
QWEN_MODEL_NAME=qwen3-maxOptional:
QWEN_BASE_URL=https://dashscope.aliyuncs.com/compatible-mode/v1
TWITTER_CLI_BIN=twitter
XHS_CLI_BIN=xhs
BILI_BIN=bili
ASR_MODEL=Qwen/Qwen3-ASR-1.7B
DEEPFIND_AUDIO_DIR=audio
DEEPFIND_TOOL_TIMEOUT=90uv run -m deepfind.cli "小红书博主:刘小鸭的AI日记 发过哪些内容?粉丝数?" --num-agent 2
uv run -m deepfind.cli "帮我总结https://www.bilibili.com/video/BV1tew5zVEDf 关于话题世界模型、逃出硅谷、反OpenAI 、AMI Labs、两次拒绝Ilya、杨立昆、李飞飞和42的内容" --num-agent 1
uv run -m deepfind.cli "same query" --num-agent 2 --quiet
uv run -m deepfind.cli "same query" Flags:
query: required--num-agent:1..4--max-iter-per-agent: default50--quiet: disable formatted progress output
- Lead agent splits the query into a few tasks.
- Sub-agents call local tools such as
xhs_search_user,xhs_user,xhs_user_posts,xhs_read,twitter_search,twitter_read, andbili_transcribe. - Lead agent merges the results into one answer.
Qwen is used through the OpenAI-compatible chat.completions API.
bili_transcribe is available to sub-agents and accepts either a Bilibili video URL
or a raw BV... ID.
It returns transcript text only (no summary generation).
If audio/transcripts/<BVID>.txt already exists,
the tool reuses it and skips download + ASR transcription.
Setup (WSL):
uv tool install bilibili-cli
bili statusArtifacts:
- Segments:
audio/<BVID>/seg_* - Transcript:
audio/transcripts/<BVID>.txt
python3 -m unittest discover -s tests -v