What is this?
Pickoo! is a local-first AI favorites workbench where a connected agent automatically ingests links you paste or share (from Xiaohongshu, Douyin, Bilibili, WeChat, Xiaoyuzhou podcasts, web pages, NetEase Music) and files them into editable Markdown with a preserved evidence package.
Pickoo! is a local-first AI favorites workbench where a connected agent automatically ingests links you paste or share (from Xiaohongshu, Douyin, Bilibili, WeChat, Xiaoyuzhou podcasts, web pages, NetEase Music) and files them into editable Markdown with a preserved evidence package.Is this a browser extension or a standalone local app?It is a standalone local service with its own full browser UI (a desktop-style card view), not a browser extension; the data layer is just editable Markdown plus an evidence package, not tied to any single platform.
It is a standalone local service with its own full browser UI (a desktop-style card view), not a browser extension; the data layer is just editable Markdown plus an evidence package, not tied to any single platform.How do I get it running?Double-click start.command on macOS or start-windows.cmd on Windows after installing the prerequisites, which opens the local service and browser UI together.
Double-click start.command on macOS or start-windows.cmd on Windows after installing the prerequisites, which opens the local service and browser UI together.Which AI agents can connect to Pickoo?Codex, WorkBuddy, Hermes, Claude and Claudian can all connect and share the same library and task state, so you can keep searching, organizing or writing from any of them instead of re-uploading material between tools.
Codex, WorkBuddy, Hermes, Claude and Claudian can all connect and share the same library and task state, so you can keep searching, organizing or writing from any of them instead of re-uploading material between tools.Does it bypass platform logins, DRM, or region locks to grab content?No. Platform login walls, memberships, DRM, region and access restrictions still apply, and anything that cannot be fully captured is explicitly shown as pending review, partially completed, or failed.
No. Platform login walls, memberships, DRM, region and access restrictions still apply, and anything that cannot be fully captured is explicitly shown as pending review, partially completed, or failed.Does the local OCR/ASR setup work on Windows and Linux too, or just macOS?Yes: Apple Silicon Macs get MLX Whisper, Intel Macs and Windows x64 get Faster-Whisper CPU int8, and Windows/Linux get RapidOCR with ONNX Runtime for OCR, all written to the same evidence format.
Yes: Apple Silicon Macs get MLX Whisper, Intel Macs and Windows x64 get Faster-Whisper CPU int8, and Windows/Linux get RapidOCR with ONNX Runtime for OCR, all written to the same evidence format.Where does my saved content and login data actually live?Everything is written to a local library on your machine (default ~/Documents/Pickoo Library), settings live in ~/.config/pickoo/settings.json, and the NetEase Music login session is stored in a private local file that never enters shared settings or API responses; the backend only listens on 127.0.0.1.
Everything is written to a local library on your machine (default ~/Documents/Pickoo Library), settings live in ~/.config/pickoo/settings.json, and the NetEase Music login session is stored in a private local file that never enters shared settings or API responses; the backend only listens on 127.0.0.1.What content sources does Pickoo support?Xiaohongshu posts, videos and public collections, WeChat official-account articles, Douyin, Bilibili, generic web pages, Xiaoyuzhou podcasts, and NetEase Music track metadata (with optional authenticated audio when you log in through the settings page).
Xiaohongshu posts, videos and public collections, WeChat official-account articles, Douyin, Bilibili, generic web pages, Xiaoyuzhou podcasts, and NetEase Music track metadata (with optional authenticated audio when you log in through the settings page).What are the six asset types it extracts from a source?Each saved item can be broken down into six reusable asset categories: quotes, viewpoints, hooks, structure notes, images, and video shots, but only candidates that pass an evidence-review quality gate are admitted into the asset library.
Each saved item can be broken down into six reusable asset categories: quotes, viewpoints, hooks, structure notes, images, and video shots, but only candidates that pass an evidence-review quality gate are admitted into the asset library.Why does it keep the raw transcript instead of just an AI summary?Pickoo stores the original text, images, audio/video, source metadata and derived OCR/ASR/chapter/asset data in separate layers linked back to the exact passage, image, timestamp or shot, so an agent can verify a conclusion against its original evidence instead of trusting a summary.
Pickoo stores the original text, images, audio/video, source metadata and derived OCR/ASR/chapter/asset data in separate layers linked back to the exact passage, image, timestamp or shot, so an agent can verify a conclusion against its original evidence instead of trusting a summary.What is the RED Skill package mentioned in the README?It is a trimmed, platform-compliant distribution of the frontend, local service and three Agent Skills, generated by npm run release:redskill, that drops large custom fonts and GitHub branding assets to stay under the 10MB-per-file and 30MB-total size limits required by that platform.
It is a trimmed, platform-compliant distribution of the frontend, local service and three Agent Skills, generated by npm run release:redskill, that drops large custom fonts and GitHub branding assets to stay under the 10MB-per-file and 30MB-total size limits required by that platform.What are similar projects?chubbyskills, autoclip, WeChatBridge, second-brain-os, qiaomu-clipper, claude-obsidian, obsidian-second-brain, VidBee
chubbyskills, autoclip, WeChatBridge, second-brain-os, qiaomu-clipper, claude-obsidian, obsidian-second-brain, VidBeeWhich topics is it in?speaker-diarization, podcast, douyin, xiaohongshu, personal-knowledge-management, yt-dlp, bilibili, obsidian, transcription, speech-to-text, note-taking, wechat
speaker-diarization, podcast, douyin, xiaohongshu, personal-knowledge-management, yt-dlp, bilibili, obsidian, transcription, speech-to-text, note-taking, wechat