1. Conduid
  2. Developer Tools
  3. Mittdigitalaminne
MCP server · Developer Tools

Mittdigitalaminne

Personal knowledge management app with MCP integration for Claude AI. Swedish UI.

Unclaimed devtools
34Low

Scored 7 months ago · breakdown

About Mittdigitalaminne

Mittdigitalaminne is an MCP server published by Pluggentipsar in the Developer Tools category: personal knowledge management app with MCP integration for Claude AI. Swedish UI.

The repository has 7 stars and 2 forks, with the last commit 5 months ago. Recent commit activity is the strongest signal that reported issues will get attention.

Install

From source
git clone https://github.com/Pluggentipsar/mittdigitalaminne

This server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.

Tools (10)

add_memorySave a new memory (auto-detects YouTube/LinkedIn/Instagram from URL)

Parameters not published.

add_noteAdd a comment/note to a memory

Parameters not published.

delete_memoryDelete a memory

Parameters not published.

get_memoryGet a specific memory by ID

Parameters not published.

get_notesList all notes for a memory

Parameters not published.

get_statisticsDashboard statistics (counts by type, recent activity)

Parameters not published.

list_tagsList all tags with usage counts

Parameters not published.

search_memoriesFull-text search with filters (type, tags, dates, favorites)

Parameters not published.

update_memoryUpdate title, summary, content, tags, favorite status

Parameters not published.

upload_imageUpload a base64-encoded image

Parameters not published.

Ask AI

✦

Ask AI about Mittdigitalaminne

Powered by Claude · Grounded in docs

I know everything about Mittdigitalaminne. Ask me about installation, configuration, usage, or troubleshooting.

Security checks

  • ✓README presentThe repository ships documentation.

5 checks have not been run yet.

README

Mitt Digitala Minne

A personal knowledge management app where you save, organize, and search memories — thoughts, links, articles, images, YouTube videos, and social media posts. Features a web UI and an MCP server for saving memories directly via Claude AI.

Swedish UI · Full-text search · AI-powered via MCP · Self-hosted


Architecture

mittdigitalaminne/
├── web/                    # Next.js web app (frontend + API)
├── mcp-server/             # MCP server (Claude AI integration)
├── supabase/
│   └── migrations/         # Database schema (run in order)
└── README.md
Component Tech Deploys to
Web app Next.js 16, React 19, TailwindCSS 4, Supabase SSR Vercel
MCP server Node.js, @modelcontextprotocol/sdk, dual transport Cloudflare Workers (HTTP) + local (stdio)
Database PostgreSQL via Supabase, Swedish FTS, trigram search Supabase

How it works

  • Web UI — Browse, search, filter, create, and edit memories at your-app.vercel.app
  • MCP server — Claude AI can save/search/manage memories via tool calls
    • stdio transport — for Claude Desktop (local)
    • HTTP transport — for Claude.ai via Cloudflare Workers (remote)

Prerequisites


Setup Guide

1. Clone and install

git clone https://github.com/Pluggentipsar/mittdigitalaminne.git
cd mittdigitalaminne

# Install dependencies for both packages
cd web && npm install && cd ..
cd mcp-server && npm install && cd ..

2. Create Supabase project

  1. Go to supabase.com/dashboard and create a new project
  2. Note your Project URL and API keys from Settings > API
  3. Run the database migrations in order via the SQL Editor:
supabase/migrations/001_initial_schema.sql
supabase/migrations/002_improved_search.sql
supabase/migrations/003_memory_notes.sql
supabase/migrations/004_new_content_types.sql
supabase/migrations/005_smart_spaces.sql
supabase/migrations/006_inbox.sql
supabase/migrations/007_projects.sql
supabase/migrations/008_snapshots.sql
  1. Create a storage bucket for images:
    • Go to Storage in the Supabase dashboard
    • Create a new bucket called memory-images
    • Set it to Public

3. Configure environment variables

Web app — create web/.env.local:

NEXT_PUBLIC_SUPABASE_URL=https://YOUR_PROJECT.supabase.co
NEXT_PUBLIC_SUPABASE_ANON_KEY=your_anon_key_here
SUPABASE_SERVICE_ROLE_KEY=your_service_role_key_here
OPENAI_API_KEY=your_openai_key_here           # For AI features (summarize, tags, etc.)

MCP server — create mcp-server/.env:

SUPABASE_URL=https://YOUR_PROJECT.supabase.co
SUPABASE_SERVICE_ROLE_KEY=your_service_role_key_here

You'll find these values in your Supabase project under Settings > API.

4. Run locally

# Terminal 1: Web app
cd web
npm run dev
# Opens at http://localhost:3000

# Terminal 2: MCP server (for testing)
cd mcp-server
npm run dev

Deployment

Deploy web app to Vercel

cd web
npx vercel --prod

Set these environment variables in your Vercel project settings:

Variable Value
NEXT_PUBLIC_SUPABASE_URL https://YOUR_PROJECT.supabase.co
NEXT_PUBLIC_SUPABASE_ANON_KEY Your anon/public key
SUPABASE_SERVICE_ROLE_KEY Your service role key
OPENAI_API_KEY Your OpenAI API key

Deploy MCP server to Cloudflare Workers

cd mcp-server
npm run build

# Set secrets (you'll be prompted for values)
wrangler secret put SUPABASE_URL
wrangler secret put SUPABASE_SERVICE_ROLE_KEY
wrangler secret put MCP_AUTH_TOKEN       # Choose a strong random token

# Deploy
npm run worker:deploy

Your MCP server will be available at https://your-worker.your-subdomain.workers.dev.


Connect Claude AI

Option A: Claude Desktop (local, stdio)

Add to your Claude Desktop config (claude_desktop_config.json):

{
  "mcpServers": {
    "mitt-digitala-minne": {
      "command": "node",
      "args": ["C:/path/to/mittdigitalaminne/mcp-server/dist/index.js"],
      "env": {
        "SUPABASE_URL": "https://YOUR_PROJECT.supabase.co",
        "SUPABASE_SERVICE_ROLE_KEY": "your_service_role_key"
      }
    }
  }
}

Build first with cd mcp-server && npm run build.

Option B: Claude.ai (remote, HTTP via Cloudflare Workers)

  1. Go to claude.ai/settings/integrations
  2. Add a new MCP integration:
    • URL: https://your-worker.your-subdomain.workers.dev/mcp
    • Authentication: Bearer token (the MCP_AUTH_TOKEN you set)
  3. Claude.ai can now save memories, search, add notes, etc.

MCP Tools

The MCP server exposes 10 tools that Claude can use:

Tool Description
add_memory Save a new memory (auto-detects YouTube/LinkedIn/Instagram from URL)
search_memories Full-text search with filters (type, tags, dates, favorites)
get_memory Get a specific memory by ID
update_memory Update title, summary, content, tags, favorite status
delete_memory Delete a memory
list_tags List all tags with usage counts
get_statistics Dashboard statistics (counts by type, recent activity)
upload_image Upload a base64-encoded image
add_note Add a comment/note to a memory
get_notes List all notes for a memory

Features

Content Types

  • 7 content types: Thought, Link, Article, Image, YouTube, LinkedIn, Instagram
  • Auto-detection of platform from pasted URLs

Rich Previews

  • YouTube — embedded video player with thumbnail and play button overlay
  • Instagram — full embed of posts/reels with dynamic height via postMessage
  • LinkedIn — branded text preview with platform colors
  • Articles & Links — OG image previews with favicon and domain
  • Smart title handling — auto-truncation of long social media titles at ~120 chars

Search & Filtering

  • Swedish full-text search with relevance ranking (weighted: title > summary > content)
  • Filters — by content type, tags, project, date range, favorites, inbox status
  • Smart Spaces — save any active filter combination as a named space in the sidebar

Organization

  • Tags with colors for categorization
  • Projects & collections — group related memories with name, color, icon, deadline, and status (active/archived)
  • Inbox — new memories land in inbox until organized
  • Favorites — star important memories for quick access

AI Features

  • AI summarization — automatic summary of articles and posts
  • AI tag suggestions — keyword extraction matched against existing tags
  • AI actions panel — process memories with AI directly in the detail view (summarize, extract key points, translate, etc.)
  • Related memories — automatic connections between similar memories based on shared tags

Web Snapshots

  • Save HTML snapshots of web pages for offline reading
  • View snapshots in a sandboxed iframe within the app

Export & Download

  • Download as .txt — clean text file with title, summary, content, metadata, and tags
  • Download as .docx — formatted Word document with typography, headings, bullet lists, and metadata
  • Available from both the context menu (right-click on cards) and the detail view toolbar
  • Works on mobile via long-press

Context Menu (Desktop & Mobile)

  • Right-click (desktop) or long-press (mobile, 500ms) on memory cards:
    • Toggle favorite
    • Add to project (submenu with active projects)
    • Download as .txt
    • Download as .docx
    • Open original link
    • Delete memory
  • Haptic feedback on mobile (vibration API)
  • Viewport-aware positioning (menu stays on screen)

Dashboard

  • Statistics overview with type distribution chart
  • Recent memories list
  • Resurfacing — rediscover old memories that may be relevant again

Comment Threads

  • Add notes/comments to any memory
  • Threaded discussion per memory

Design

  • "Amber Glow" theme — warm amber accents with content-type colored card glows
  • Fonts: Instrument Serif (headings) + DM Sans (body text)
  • Collapsible sidebar (72px collapsed, 272px expanded) with localStorage persistence
  • Responsive — works on desktop and mobile
  • Ambient gradient orbs and decorative dividers
  • Swedish UI throughout

Project Structure

web/
├── src/
│   ├── app/
│   │   ├── api/                # API routes
│   │   │   ├── memories/       # CRUD + notes + related
│   │   │   ├── projects/       # Projects CRUD + member management
│   │   │   ├── spaces/         # Smart Spaces CRUD
│   │   │   ├── suggest-tags/   # AI tag suggestions
│   │   │   ├── resurface/      # Memory resurfacing
│   │   │   ├── unfurl/         # URL metadata + article extraction
│   │   │   ├── statistics/     # Dashboard stats
│   │   │   ├── tags/           # Tag management
│   │   │   └── upload/         # Image upload
│   │   ├── dashboard/          # Dashboard page
│   │   ├── inkorg/             # Inbox page
│   │   ├── lagg-till/          # Create memory page
│   │   ├── minnen/             # Memory list + detail + edit pages
│   │   ├── projekt/            # Project list + detail pages
│   │   └── taggar/             # Tag management page
│   ├── components/
│   │   ├── memories/           # MemoryCard, MemoryGrid, CardContextMenu, AiActionsPanel
│   │   │   └── previews/       # YouTubePreview, InstagramEmbed, SocialPreview, LinkPreview
│   │   ├── projects/           # AddToProjectDialog, ProjectCard
│   │   ├── spaces/             # SaveSpaceDialog
│   │   ├── dashboard/          # Stats, charts, recent list, resurfacing
│   │   ├── filters/            # FilterBar (type, tags, project, search, save space)
│   │   ├── layout/             # Sidebar navigation + spaces
│   │   ├── contexts/           # SidebarContext
│   │   └── ui/                 # MarkdownContent, SnapshotViewer
│   ├── hooks/                  # SWR data hooks (memories, tags, spaces, projects, suggestions)
│   └── lib/                    # Supabase clients, types, utils, export utilities, stopwords

mcp-server/
├── src/
│   ├── server.ts               # MCP server factory (shared logic)
│   ├── index.ts                # stdio entry point (Claude Desktop)
│   ├── worker.ts               # Cloudflare Workers entry (Claude.ai)
│   ├── tools/                  # Individual tool handlers
│   ├── utils/                  # URL detection, tag helpers
│   └── db/                     # Supabase client factory
├── wrangler.toml               # Cloudflare Workers config
└── tsconfig.json

supabase/
└── migrations/
    ├── 001_initial_schema.sql          # Core tables, FTS, RLS
    ├── 002_improved_search.sql         # Enhanced search with tags in FTS
    ├── 003_memory_notes.sql            # Comment/notes table
    ├── 004_new_content_types.sql       # YouTube, LinkedIn, Instagram types
    ├── 005_smart_spaces.sql            # Smart Spaces (saved filters)
    ├── 006_inbox.sql                   # Inbox flag for memories
    ├── 007_projects.sql                # Projects & memory_projects tables
    └── 008_snapshots.sql               # Snapshot HTML storage

License

MIT

README copy captured when this server was indexed; it may lag the repository. The original is authoritative.

Questions

About Mittdigitalaminne

How do I install Mittdigitalaminne?

No package has been published for this server, so install it from source: clone the repository and follow its README.

Is Mittdigitalaminne safe to use with an AI agent?

Its trust score is 34 out of 100 (low). It passes 1 of the 1 static security checks run so far. It has no ConduID identity yet, so agent calls to it are not receipted.

Is Mittdigitalaminne still maintained?

Yes — the last commit was 5 months ago. The repository has 7 stars and 2 forks.