AI Ingest
Research anything
Ask your AI to explore a topic, explain a concept, or analyse an article. Then generate all 10 knowledge angles from the conversation.
Studio
No angles generated yet
Generate 10 knowledge angles from any content — paste text, chat with AI, or pick an article from RSS.
RSS Ingest
Articles
Feeds
Working
Paywalled
Feeds
Fetches via your browser + rss2json — zero AI tokens. Hover a feed's status dot to see which path served it.
Refresh feeds to load articles
URL Ingest
Any article, blog, or news URL
Substack, Medium, Dev.to, GitHub
Works on most public articles
Paywalled sites — fetch article, then paste text
Send any page here in one click: 📥 Send to Studio ← drag to your bookmarks bar
Publish Tracker
Filter:
No articles tracked yet. Generate and save content — it appears here automatically.
Library
Pick an article on the left to read it.
Discover

Pulls one random article out of your saved library and opens it in the reader — old research doesn't do you any good if it just sits there.

Flashcards
Digest
Pick an article on the left to see its top snippets.
Report/Telemetry
No reading data yet
Generate content and read through the angles. Your reading behaviour, time spent, and save patterns are tracked here.
Help
Getting started
Core workflow
Exports & integrations
Settings & account
Reference
Settings
AI Provider
Select your AI model for all generation
Not set
Stored locally. console.anthropic.com ↗
Obsidian Vault PRIMARY
Save notes directly — 10 angle files + index + canvas + daily note + Kanban
Not set
Chrome & Edge only (File System API). Files are plain .md — your data is safe forever.
What gets saved:
  1. Publish/ · KB-Notes/ · Interview-Prep/ · Study-Guides/ · AI-View/ · Frameworks/ · Compliance/
  2. topic-index.md — links all 10 angles + Reading Summary
  3. _domains/ — domain hub notes for graph view
  4. topic.canvas — visual map of all 10 angles
  5. Journal/YYYY-MM-DD.md — daily note entry
  6. _tracker/kanban.md — Publish Tracker Kanban board
Fetch Interfaces
Backend + helper proxies for RSS Fetch All
Not set
Node.js Backend — RSS — optional
Your own Node server — if set, RSS Fetch All tries this first, skipping public CORS proxies entirely. Leave blank and it fetches straight from your browser instead (direct fetch, then the helper proxies below) — no setup required either way. API key is required so strangers can't use your server as a free proxy — must match your server's RSS_API_KEY env var.
Python Backend — RSS — optional, tried after the Node.js Backend — RSS above
A separate Python app (server/py-rss-app, deployed via cPanel's "Setup Python App") using feedparser — its own deploy, own port, own key, fully independent of the Scrape one below (mirrors how the Node.js RSS/Scrape backends are already split). Tried only if the Node.js Backend — RSS above isn't configured or fails, before falling back to your browser/public proxies.
Node.js Backend — Scrape — optional, separate server
A separate Node server (server/scrape-app) — deliberately its own app, own port, own key, not shared with the RSS backend above. If set, URL Ingest and "Fetch Full Article" try this first, before the public proxy chain. Requests from your own server carry less abuse reputation than the shared public proxies, so Cloudflare-fronted sites (Engadget, The Verge, etc.) are more likely to let it through.
Python Backend — Scrape — optional, tried after the Node.js Backend — Scrape above
Same Python server, using curl_cffi to impersonate a real browser's TLS fingerprint for article pages — genuinely different from what the Node backends can do. Tried only if the Scrape Backend above isn't configured or fails, before the public proxy chain. Independent of the RSS one above — set only this if you just want the fallback for article scraping.
NewsAPI.org — helper
Free: 100 req/day · 80,000+ sources. newsapi.org ↗
rss2json.com — helper, optional
RSS Fetch All uses rss2json's shared free-tier quota, which can 429 under load from other anonymous users. A personal key gives you your own quota. Free tier: 10,000 req/day. rss2json.com ↗
Exports
Google Drive · Notion · OneNote — click to expand
Not set
Google Drive
Not connected
◼ Notion
Not set
OneNote — online, cross-platform
Not connected
Setup guide ↗ — works on any browser, including Safari
Logseq — local graph, Chrome/Edge only
Not connected
Writes pages into pages/ and a linking bullet into today's journals/ entry, in your Logseq graph folder — same local File System Access API as the Obsidian vault above, so it needs its own folder pick and only works in Chrome/Edge.
🧠
Your Context
Tells the AI who you are and what to prioritize — steers framing/emphasis, never overrides the source article's facts
Not set
Stored locally only. Leave blank to keep responses generic.
🔖
Quick Ingest Bookmarklet
Drag to your bookmarks bar — click it on any article to send it here
Works in every browser — it's just a bookmark, no extension install needed.

Think with AI. Remember with Obsidian. TheVaultMind generates 10 structured knowledge formats from any content and saves them directly to your Obsidian vault as plain markdown. Your notes live on your disk — nothing in the cloud, no account required, no one else can see your research.

2026 Sarbjit Singh & Gurtej Singh
Signed in as
Session on this browser only — logging out doesn't affect other devices