Rakenne

Release 0.4.0 adds in-app editing of output documents with a rich-text editor, plus private/public visibility for skills and submit-for-review for skills with custom code.

Edit documents in the app

Output files from the agent are no longer read-only. When you’re viewing a document in the preview panel, use Edit in the toolbar to open it in a full rich-text editor in a new window.

• Format as you type — Headings, bold, italic, underline, lists, blockquotes, code blocks, links, and tables. The toolbar gives you one-click formatting and the editor keeps everything in markdown under the hood.

• Save back to the project — Your changes are written to the workspace file. The preview in the main window refreshes when you save, and exports (Word, PDF, etc.) use the updated content.

• No round-trip to the agent — Fix typos, tweak wording, or adjust structure yourself instead of asking the agent every time.

The editor opens in a separate window so you can keep the chat and preview in place. You’ll get a warning if you try to leave with unsaved changes.

Private and public skills

Your organization’s skills now have a visibility setting:

• Private (default) — Only your organization can see and install the skill. Shown with a lock icon in the library and next to installed skills in the workspace.

• Public — Visible to all Rakenne users. Shown with a globe icon.

You can switch a skill between private and public from the skill detail dialog. Markdown-only skills can be made public anytime. Skills that include custom code (e.g. TypeScript or JavaScript) must be submitted for review first; once approved, you can set them to public. The library shows badges (e.g. “Pending review”, “Approved for public”) so you know the status at a glance.

Organization slug in settings

In Settings → General, tenant admins can view and edit the organization’s public slug — the short identifier used when sharing your public skills. It’s generated from your organization name when the tenant is created and can be changed if you need a different one.

Hi! I’m the creator of Rakenne. I built this because I noticed a recurring problem with LLMs in professional settings: chat-based document creation is unpredictable and hard to scale for domain experts.

Experts know the process of building a document (the questions to ask, the order of operations, the edge cases), but translating that into a long system prompt often leads to hallucinations or missed steps.

What is Rakenne? Rakenne is a multi-tenant SaaS that lets domain experts define "Guided Workflows" in Markdown. An LLM agent then runs these workflows server-side, conducting a structured dialogue with the user to produce a final, high-fidelity document.

The Tech Stack:

• Agentic Core: Built on the pi coding agent using RPC mode. This allows the agent to maintain state and follow complex logic branches defined in the Markdown files.

• Frontend: Built with Lit web components. I wanted something incredibly lightweight and framework-agnostic so the document "interviews" feel snappy and can eventually be embedded as widgets.

• Multi-tenancy: Designed to isolate agent environments server-side, ensuring that custom expert logic doesn't leak between tenants.

Why this approach? Instead of "Chat with a PDF," it’s "The Logic of an Expert." If you’re a lawyer or a compliance officer, you don’t want a creative partner; you want a system that follows your proven methodology. By using Markdown, we make the "expert logic" version-controllable and easy for non-devs to edit.

I’d love your feedback on:

1. The Agentic UX: Does the "interview" flow feel natural, or is it too rigid?

2. Markdown as Logic: Is Markdown the right "DSL" for this, or should we move toward something like YAML or a custom schema?

3. Latency: We're using RPC for the agent-browser communication—is the response time acceptable for your use case?

Demo (No signup required): https://rakenne.app

Thanks! I'll be around to answer any technical questions.