Jamailar/Beav
小红书 AI 运营工作台|小红书采集、评论区下载、素材库、选题、AI写作、小红书全域解决方案,开箱即用,一键安装,小红书AI工作台,自媒体素材库,AI写作+图片自动编排,小红书版OpenClaw、自媒体资产底座+AI工作台,支持小红书图文+小红书评论区下载、小红书AI创作工具、自媒体版WorkBuddy、抖音、小红书爬虫数据采集
Beav – AI‑powered content‑creation workbench
What it is
- A desktop application (macOS, Windows, Linux) that bundles a knowledge base, asset library, AI‑assisted writing, image generation, video editing and publishing tools into one “workspace” for social‑media creators.
- It keeps all data locally, adds long‑term memory per account/brand, and can hook into external AI agents such as Claude, Codex, WorkBuddy, etc.
Core capabilities
| Area | What you can do |
|---|---|
| Content collection | Paste a link (or upload media) from your phone, have Beav download the video, transcribe it, and automatically break it into topic, structure, shot list and raw script. Subscribing to rival creators lets you sync new posts daily. |
| Planning | Daily “运营日报” (operations report) at 09:00, hotspot ranking that scores trending topics for your account, and a calendar view (month/week/year) for scheduling posts. |
| Writing | In a chat‑style pane, tell the AI what you need (e.g., “write a Xiaohongshu note about autumn shoes”). The AI reads your knowledge base and account profile, then outputs a ready‑to‑publish manuscript with title, body, tags and optional images, saved as a material package. |
| Design & video | Template library for cover images and graphics; AI can generate images from prompts. Smart‑cut for spoken‑word videos: the AI transcribes, suggests cuts (remove filler, repeats), adds subtitles, animated text, picture‑in‑picture and other effects. Full multi‑track video editor with keyframes, transitions, colour grading, audio denoise, and export of video or separate SRT/WebVTT files. |
| Publishing | Material packages are QR‑coded; scan with your phone to pull the original media, copy the caption and post directly on the target platform. |
| Agent integration | One‑click connectors for popular coding‑assistant tools (Claude Code, Codex, WorkBuddy, Cursor). Those agents can read/write Beav’s libraries, launch creation tasks, and retrieve results, letting you keep Beav as a “knowledge‑aware asset store” while you work in your favourite IDE. |
| Mini‑apps & skill market | Built‑in utilities (batch caption extraction, link download, subtitle generation, AI voice‑over, image‑to‑Live‑photo, etc.) and a marketplace where you can install additional AI‑driven skills for specific workflow steps. |
Technology stack
- UI & cross‑platform packaging: Tauri (Rust + web front‑end).
- Core AI integration: configurable OpenAI‑compatible endpoints or local model servers; also ships with a default official AI service.
- Local storage: files on disk, with optional links to external folders, Obsidian vaults or Eagle asset manager.
- Video processing: native Rust libraries for transcoding, speech‑to‑text, and timeline editing.
License
- MIT‑NC (non‑commercial). Commercial use requires the author’s written permission.
Getting started
- Download the installer for your OS from the official site or the GitHub Releases page.
- Create a space for each account/brand; paste the account’s URL so the AI can initialise a profile.
- Add content via the mobile‑to‑desktop QR flow or the Chrome/Edge extension.
- Use the daily report or hotspot view to pick a topic, then start a new chat and ask the AI to write, generate images, or edit video.
- When the material package is ready, scan the QR code to pull it onto your phone and publish.
Who it’s for
- Solo creators managing one or several social‑media accounts.
- Video‑essay or knowledge‑sharing creators who need quick cuts and subtitles.
- Writers of image‑heavy posts (Xiaohongshu, WeChat public accounts, newsletters).
- Small agencies or teams that want a shared, locally stored asset library with AI‑enhanced workflows.
Community & support
- Official website & download portal: https://beav.pro/
- Bilibili tutorial series, GitHub Issues for bug reports, and a roadmap/CHANGELOG in the repo.
- Business and reseller contacts are listed in the README.
All information above is taken directly from the repository’s README; no additional features have been inferred.
Related
- Project
- Project
- Project
- Project
- Project