# Hjarni for Developers Note-taking for developers, built so your AI can actually read it. Stop re-explaining your stack every time you open a new chat. 25 notes free, forever. No credit card required. We’ll keep the template ready while you connect your assistant. ## Ask about your own architecture, not a generic stack When your runbooks, architecture decisions, and team conventions live in Hjarni, Claude and ChatGPT can answer with the context your project actually uses. Claude answers a project-specific engineering question using notes stored in Hjarni. ## The problem Every conversation with Claude or ChatGPT starts from scratch on your codebase. You explain your project structure. Your naming conventions. Why you chose Postgres over MySQL. That the auth layer uses a custom middleware. Then you do it again tomorrow. ## Note-taking for developers, built for your AI to read Write down what your AI needs to know. Architecture decisions. API conventions. Deployment notes. Debugging playbooks. Connect Hjarni as an MCP server and your AI can search and read those notes whenever a task calls for them, or when you ask. A standing instruction to check Hjarni first makes it look before it answers. A typical developer setup - **Architecture folder**: stack overview, service boundaries, database schema notes - **Conventions folder**: naming, error handling, test patterns, PR guidelines - **Debugging folder**: known issues, past incidents, environment quirks - **AI instructions on each folder**: "When asked about deployments, check the runbook first" ![The folder sidebar in Hjarni showing Architecture, Conventions, and Debugging folders, each with one note.](https://hjarni.com/use-cases/developers.png) ## Skip the blank slate The [Knowledge Wiki template](https://hjarni.com/templates/knowledge-wiki) gives you a ready-made structure for capturing architecture decisions, conventions, and operational knowledge. Paste the template link into Claude or ChatGPT and it creates the initial structure for you. ## A concrete workflow You're debugging a production issue. You ask Claude about the retry logic in your payment service. Claude checks your notes and finds you use Sidekiq with exponential backoff. That retries are capped at 5. That the payment service wraps Stripe. It doesn't guess. It reads your notes. After the fix, you save what you learned. Next conversation, it's already there. Questions like "What does our auth middleware do?" and "Why did we choose Sidekiq?" get better answers when the reasoning is already written down and your assistant can read it. ## The gist stack, hosted Karpathy's LLM Wiki gist showed the pattern. A maintained brain instead of RAG. Run it locally if you live in a terminal. Hjarni is the same pattern, hosted. Claude and ChatGPT read your notes. No vault on your laptop. The full breakdown is in [Karpathy's LLM Wiki is right](https://hjarni.com/blog/karpathys-llm-wiki-is-right). ## Why not just use a README or wiki? READMEs are for humans. Your AI doesn't read your GitHub wiki when you open a new chat. Hjarni is a note system that both you and your AI can use. Folder-level instructions shape how the AI behaves in different contexts. > Your AI remembers a few facts about you. It doesn't remember your codebase. ## What developers keep in Hjarni - **Architecture notes**: service boundaries, design decisions, and system tradeoffs - **Operational runbooks**: deploy steps, incident notes, rollback plans, and environment quirks - **Engineering conventions**: naming, testing, error handling, and review expectations - **Project-specific instructions**: folder-level guidance that tells your assistant how to behave in each codebase - **Team memory**: a shared place for the context new engineers and new AI chats always need - **Portable notes**: Markdown export and API access when you want to use the data elsewhere ## When your whole team needs it The notes that keep your AI current can become your team's shared context. Put architecture, conventions, and runbooks in a shared team space and every engineer's Claude or ChatGPT reads the same source. Per-folder roles let you decide who can edit what, and a new hire can ask from that context on day one instead of interrupting a senior engineer. See [Hjarni for Teams](https://hjarni.com/for-teams). ## Skip the blank page Choose a template, connect your assistant, and ask it to create the folders and instructions. Then save a real note and ask for it in a new chat. ### [Developer Project Memory](https://hjarni.com/templates/developer-project-memory) A ready-made knowledge base for your codebase. Architecture, conventions, decisions, runbooks. Claude and ChatGPT read it before they write a line. ### [Knowledge Wiki](https://hjarni.com/templates/knowledge-wiki) Drop articles, links, and ideas here. Your AI files them, keeps topic pages up to date, and flags contradictions, so your knowledge compounds instead of scattering across chat histories. Common questions ## Questions developers actually ask **How do I give ChatGPT context about my codebase?** Store architecture notes, conventions, runbooks, and debugging context in Hjarni, then connect ChatGPT or Claude via MCP. Your assistant can read that context at the start of the conversation instead of relying on whatever you paste into the prompt. **What should developers put in Hjarni?** The highest-value notes are architecture overviews, service boundaries, deployment runbooks, debugging playbooks, coding conventions, and the decisions behind past tradeoffs. Those are exactly the details your AI usually lacks. **Can I use Hjarni as an AI coding assistant memory?** Yes. Hjarni acts as the long-term memory layer for your existing AI coding assistant. It does not replace ChatGPT or Claude. It gives them reusable project context they can read across sessions. **Can my engineering team share AI context?** Yes. Teams can share folders for architecture, incidents, conventions, and runbooks so everyone and every connected assistant works from the same context. **Does Hjarni work for debugging and incident response?** Yes. It is a strong fit for incident notes, known failure modes, rollback steps, and environment quirks. That gives your AI useful operational memory instead of just code generation context.