Hjarni pitches MCP-native notes app as antidote to tool bloat
Developer Evert argues that over-featured tools like Notion and Obsidian waste users' time on setup rather than actual work, and introduces Hjarni — a minimal notes app with a built-in MCP server so AI assistants can read your notes directly.
Score breakdown
Developers building or choosing AI-integrated tooling should take note: exposing personal knowledge bases via MCP is emerging as a practical pattern for persistent AI context — Hjarni is an early, opinionated example of what "MCP-first" product design looks like in practice.
- 01Author Evert published the piece on Dev.to under the #mcp tag on April 19, 2026.
- 02Evert argues Notion and Obsidian impose a 'setup tax' — time spent on configuration before any actual note-taking occurs.
- 03Hjarni is described as a knowledge base built around exactly five features: notes, folders, tags, search, and a built-in MCP server.
In a pointed opinion piece on Dev.to, developer Evert makes the case that complexity is the central failure of modern knowledge tools. He singles out Notion — with its databases, kanban boards, timelines, wikis, automations, formulas, rollups, and relations — and Obsidian, which he says does "almost nothing" out of the box and pushes users into a plugin rabbit hole involving tools like `Dataview`, `Templater`, and `Calendar`. His core claim: the tool becomes the project, and users end up spending hours on configuration with zero notes written.
The MCP (Model Context Protocol) integration is positioned as the product's defining innovation — an open protocol that lets AI tools connect to external data sources.
His solution is Hjarni, scoped to exactly five features: notes, folders, tags, search, and a built-in MCP server. The MCP (Model Context Protocol) integration is positioned as the product's defining innovation — an open protocol that lets AI tools connect to external data sources. Because Hjarni exposes notes via MCP, a user can connect once and have assistants like Claude or ChatGPT read stored context automatically across conversations, removing the need to re-explain projects, tech stacks, or constraints at the start of every session.
Evert explicitly frames this as an architectural choice, not a feature gap — "we left features out on purpose." The absence of kanban boards, calendars, databases, and AI-generated writing is intentional. The pitch is that the right mental model for a notes app in the AI era is not "build a system" but rather "write it down and let your AI remember it for you," with MCP as the connective tissue between human knowledge capture and AI retrieval.
Key facts
- 01Author Evert published the piece on Dev.to under the #mcp tag on April 19, 2026.
- 02Evert argues Notion and Obsidian impose a 'setup tax' — time spent on configuration before any actual note-taking occurs.
- 03Hjarni is described as a knowledge base built around exactly five features: notes, folders, tags, search, and a built-in MCP server.
- 04MCP (Model Context Protocol) is described as 'an open protocol that lets AI tools connect to external data.'
- 05The MCP integration is designed so AI assistants like Claude or ChatGPT can read a user's notes directly, eliminating manual context-pasting.
- 06Hjarni deliberately excludes kanban boards, calendars, databases, spreadsheet views, and AI note-writing features.
- 07Evert frames the product as 'AI-native, not AI-added,' designed from the ground up around the idea that users write and AI reads.
Topics
Summary and scoring are generated automatically from the original article. We always link back to the publisher and never republish images or paywalled content. Last processed Apr 19, 2026 · 11:33 UTC. How this works →