Save anything. Ask anything. Get answers from your own knowledge — not a stranger's AI.

Live app · Contributing · Code of Conduct · Docs

---

DumpIt is an open-source, AI-powered knowledge vault. Save links, notes, and PDFs from wherever you already are — the web app, the browser extension, email, or your phone's share sheet — then ask questions in plain English and get cited answers grounded in what you actually saved, not hallucinated from general knowledge.

Table of contents

• Features
• Platform
• How RAG works
• Tech stack
• Repository structure
• Quick start
• Environment variables
• Firestore vector indexes
• Commands
• Documentation
• Contributing

Features

• Capture from anywhere
  • Links — auto-enriched titles, descriptions, and tags via server-side scraping.
  • Notes — plain-text ideas, code snippets, structured thoughts.
  • PDFs — in-memory text extraction for uploads up to 10MB.
  • Browser extension — one-click save, side-panel review, and text-selection capture (Chrome, Manifest V3).
  • Mobile share sheet — share directly into DumpIt from any Android app (PWA, Web Share Target).
  • Email-to-save — forward anything to save@dumpit.page (built; pending DNS/Resend dashboard setup on our end before it's reachable).
• AI search & cited Q&A
  • My Dump / Shared / All search scopes across your private vault and community-shared resources.
  • Answers include inline citations and source cards you can verify.
• Organization
  • Collections, tags, filtering, cursor-based pagination, duplicate detection.
  • Public profiles at /u/[username] with per-resource visibility control.
• Developer access
  • REST API secured by long-lived API keys (generate/revoke from Settings).
  • MCP server — query and save to your vault directly from Claude Desktop or Cursor.
• Infrastructure
  • Sentry error monitoring (client, server, edge).
  • Upstash Redis rate limiting on authenticated, public, and AI-query routes.
  • Dodo Payments (Merchant of Record — handles global tax compliance automatically).

Platform

Surface	What it's for
Web app	Full dashboard — search, capture, collections, profile, settings.
Browser extension	One-click save from any tab without leaving the browser.
REST API	Programmatic add/search/ask, authenticated with an API key from Settings.
MCP server	ask_vault, search_vault, save_to_vault tools for Claude Desktop / Cursor.
Mobile share sheet	Share a link straight into your vault from Android's native share menu.

How RAG works

Saving a resource creates a resources document in Firestore. AI search relies on server-side background indexing:

1. Extraction
   • Links — fetches page content and extracts readable text.
   • PDFs — parses the binary in memory via pdf-parse.
   • Notes / emails — uses the content directly.
2. Chunking & embedding — splits text into contextual chunks, generates 768-dimensional vectors via Gemini's embedding model.
3. Storage & search — stores vectors in Firestore resource_chunks, runs vector similarity search against user queries, and returns an answer with inline citations.

flowchart TD
    Input["Link / Note / PDF / Email"] --> Extract["Extract Text (Fetch / pdf-parse)"]
    Extract --> Chunk["Chunk Text"]
    Chunk --> Embed["Gemini Embedding (768d)"]
    Embed --> Store["Firestore resource_chunks"]
    Ask["User Question"] --> QueryEmbed["Gemini Query Embedding"]
    QueryEmbed --> Vector["Firestore Vector Search"]
    Store --> Vector
    Vector --> Answer["Gemini Answer with Citations"]

Loading

Tech stack

Layer	What
Framework	Next.js 14 (App Router), TypeScript
Styling	Tailwind CSS — Zinc palette, Space Grotesk + Inter
Auth	Firebase Auth (Google sign-in only)
Database	Firebase Admin SDK + Firestore (incl. native vector search)
AI	Google Gemini — embeddings & generation
Rate limiting	Upstash Redis
Error tracking	Sentry (client, server, edge)
Email	Resend (inbound save-by-email)
Payments	Dodo Payments (Merchant of Record)
Deployment	Vercel (web app), Firebase (backend/db)

Repository structure

app/                   Next.js App Router pages and API routes
  api/                 Server-side route handlers
    resources/         CRUD for saved items (links, notes, PDFs)
    ai/                Search + RAG "ask" endpoints
    webhooks/          Dodo Payments + Resend inbound-email webhooks
    settings/api-keys/ API key generate/list/revoke
  share/               Mobile share-sheet capture page
  u/[username]/        Public user profile pages
dumpit-extension/      Chrome extension (Manifest V3, separate package)
dumpit-mcp/            MCP server for Claude Desktop / Cursor (separate package)
docs/                  Internal documentation
public/                Static assets, PWA manifest
types/                 Shared TypeScript types

Quick start

git clone https://github.com/Rayan9064/dumpit.git
cd dumpit
npm install
cp .env.example .env.local
npm run dev

Open http://localhost:3000.

To try the browser extension or MCP server locally, see their own READMEs: dumpit-extension/README.md · dumpit-mcp/README.md.

Environment variables

Configure in .env.local (and in Vercel for production):

# Firebase client SDK (browser-safe)
NEXT_PUBLIC_FIREBASE_API_KEY=
NEXT_PUBLIC_FIREBASE_AUTH_DOMAIN=
NEXT_PUBLIC_FIREBASE_PROJECT_ID=
NEXT_PUBLIC_FIREBASE_STORAGE_BUCKET=
NEXT_PUBLIC_FIREBASE_MESSAGING_SENDER_ID=
NEXT_PUBLIC_FIREBASE_APP_ID=
NEXT_PUBLIC_FIREBASE_MEASUREMENT_ID=

# Firebase Admin SDK (server-only)
FIREBASE_PROJECT_ID=
FIREBASE_CLIENT_EMAIL=
FIREBASE_PRIVATE_KEY="-----BEGIN PRIVATE KEY-----\n...\n-----END PRIVATE KEY-----\n"

# Gemini AI (server-only)
GEMINI_API_KEY=
GEMINI_MODEL=gemini-2.5-flash
GEMINI_EMBEDDING_MODEL=gemini-embedding-001

# App URL & SEO
NEXT_PUBLIC_APP_URL=https://www.dumpit.page

# Rate limiting
UPSTASH_REDIS_REST_URL=
UPSTASH_REDIS_REST_TOKEN=

# Monitoring & analytics (optional)
NEXT_PUBLIC_SENTRY_DSN=
NEXT_PUBLIC_POSTHOG_KEY=
NEXT_PUBLIC_POSTHOG_HOST=

# Resend (email-to-save)
RESEND_API_KEY=
RESEND_INBOUND_WEBHOOK_SECRET=

# Dodo Payments
DODO_PAYMENTS_API_KEY=
DODO_WEBHOOK_SECRET=

Firestore vector indexes

Ask DumpIt requires Firestore vector indexes for semantic search:

# Private search index
gcloud firestore indexes composite create \
  --project=YOUR_PROJECT_ID \
  --collection-group=resource_chunks \
  --query-scope=COLLECTION \
  --field-config=order=ASCENDING,field-path=user_id \
  --field-config=vector-config='{"dimension":"768","flat": "{}"}',field-path=embedding

# Shared / All search index
gcloud firestore indexes composite create \
  --project=YOUR_PROJECT_ID \
  --collection-group=resource_chunks \
  --query-scope=COLLECTION \
  --field-config=order=ASCENDING,field-path=is_public \
  --field-config=order=ASCENDING,field-path=user_id \
  --field-config=vector-config='{"dimension":"768","flat": "{}"}',field-path=embedding

Commands

npm run dev        # Run Next.js dev server
npm run typecheck  # TypeScript type checking
npm run build      # Build production bundle
npm test           # Run Vitest unit tests

Documentation

• Deployment Guide
• System Design
• Data Model
• API Spec
• Testing Guide
• Firebase Setup

Contributing

Contributions are welcome — see CONTRIBUTING.md for setup and workflow, and CODE_OF_CONDUCT.md for community guidelines. DumpIt is MIT licensed.