Simple self-hosted LLM assistant with user-steered compounding context
I built a personal LLM assistant on Cloudflare Workers + Durable Objects. You specify a category and topic when starting a new conversation, so the backend maintains a summary for each category/topic - building up as more conversations happen under the same one. There is no complicated RAG, embeddings, or agentic magic, but the category/topic summaries system just works, and I've genuinely found it super handy and have been using it daily for my life and work. I started it to get familiar with Durable Objects and to test out this idea draft I had buried on my board. After recently…
What it does
In the maker’s words, at launch
I built a personal LLM assistant on Cloudflare Workers + Durable Objects. You specify a category and topic when starting a new conversation, so the backend maintains a summary for each category/topic - building up as more conversations happen under the same one. There is no complicated RAG, embeddings, or agentic magic, but the category/topic summaries system just works, and I've genuinely found it super handy and have been using it daily for my life and work. I started it to get familiar with Durable Objects and to test out this idea draft I had buried on my board. After recently losing my job, I started working on it more. Testing it with my friends and focusing on it being as straightforward as possible to understand, deploy and use. This week I open-sourced it for more people to try Repo: https://github.com/kol3x/pawmc Some details: - $0 cost on Cloudflare's free tier with default configuration - Everything in a singleton Durable Object - Workers AI as the default provider - OpenRouter support for more stable access and bigger models (there, you'd need to pay for tokens)
Does the same job
all alternatives →
- GFGit for LLMs – A context management interfaceOct 2025 · twigg.ai · ▲106
Hi HN, we’re Jamie and Matti, co-founders of Twigg. During our master’s we continually found the same pain points cropping up when using LLMs. The linear nature of typical LLMs interfaces - like ChatGPT and Claude - made it really easy to get lost without any easy way to visualise or navigate your project. Worst of all, none of them are well suited for long term projects. We found ourselves spending days using the same chat, only for it to eventually break. Transferring context from one chat to another is also cumbersome. We decided to build something more intuitive to the ways humans think.…
- AGAiaiai.guide: Plain-English mental model for LLM apps, tools and agentsApr 2026 · aiaiai.guide · ▲7
I’ve been building LLM tooling for a small VC fund and found myself explaining the same mental model over and over to non-technical people around me: how a stateless LLM becomes a chatbot, how tool use works, what an agent is mechanically, and why context windows shape all of it. I never found a guide that covered that full chain at the level I wanted, so I wrote one. It’s nine short chapters, each building on the last. Deliberately simplified: the goal is a useful mental model, not a textbook. Feedback, corrections, and contributions welcome: github.com/ymyke/aiaiai
- ODObject database for LLMs that persists across chats (MCP server)2025 · dry.ai · ▲9
I’d like to use LLMs for remembering all kinds of things: fitness, to-do lists, contacts, bug reports, research links, whatever. But there is no way to do that now. For example, if I find a great coding tutorial in chat, or tell it how much I ran yesterday, it forgets that when I close the chat. Even if I keep the chat history, I still need to scour through lots of messages to find the data I want. Ideally, Claude would remember all this, and I’d be able to find it later with ease. This is what my team built. It is a collaborative database you add to any LLM that supports MCP. (Claude Code,…
- GBGraph-Based Editor for LLM Workflows2024 · github.com · ▲8
Hey HN, We’re excited to share PySpur, an open-source tool that provides a graph-based interface for building, debugging, and evaluating LLM workflows. Why we built this: Before this, we built several LLM-powered applications that collectively served thousands of users. The biggest challenge we faced was ensuring reliability: making sure the workflows were robust enough to handle edge cases and deliver consistent results. In practice, achieving this reliability meant repeatedly: 1. Breaking down complex goals into simpler steps: Composing prompts, tool calls, parsing steps, and branching…
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 27d ago · cactuscompute.com
Launched alongside, August 2026
the whole month →- TL
Life & fun · 10d ago · louisabraham.github.io
- SA
Hello HN! I found that picking out plausible but diverse skin tones for my digital art and game development projects was kind of difficult, and I got curious about if there was a way to define a color space that made it easy. I've built a color picker and procedural generation algorithm based on the space as well as a bunch of other fun js features and demos throughout the page that use the equations. If you find it interesting, I have lots of explanations of how I built it and what properties the space has. The methodology might be a bit shaky, but hopefully the result is as helpful for…
Life & fun · Aug 2026 · toneyalexander.github.io
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com