Zuckerman – minimalist personal AI agent that self-edits its own code
Hi HN, I'm building Zuckerman: a personal AI agent that starts ultra-minimal and can improve itself in real time by editing its own files (code + configuration). Agents can also share useful discoveries and improvements with each other. Repo: https://github.com/zuckermanai/zuckerman The motivation is to build something dead-simple and approachable, in contrast to projects like OpenClaw, which is extremely powerful but has grown complex: heavier setup, a large codebase, skill ecosystems, and ongoing security discussions. Zuckerman flips that: 1. Starts with almost nothing…
In plain words
Zuckerman is a minimalist personal AI agent that starts with only essential features and can modify its own code and configuration files in real time. The agent's behavior, tools, and prompts are stored in plain text files that hot-reload instantly upon changes. Multiple agents can share improvements with each other, and the system supports multiple communication channels including Discord, Slack, and Telegram. It prioritizes simplicity and approachability over complex frameworks, designed for users who want a lightweight, self-improving AI assistant.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
Hi HN, I'm building Zuckerman: a personal AI agent that starts ultra-minimal and can improve itself in real time by editing its own files (code + configuration). Agents can also share useful discoveries and improvements with each other. Repo: https://github.com/zuckermanai/zuckerman The motivation is to build something dead-simple and approachable, in contrast to projects like OpenClaw, which is extremely powerful but has grown complex: heavier setup, a large codebase, skill ecosystems, and ongoing security discussions. Zuckerman flips that: 1. Starts with almost nothing (core essentials only). 2. Behavior/tools/prompts live in plain text files. 3. The agent can rewrite its own configuration and code. 4. Changes hot-reload instantly (save -> reload). 5. Agents can share improvements with others. 6. Multi-channel support (Discord/Slack/Telegram/web/voice, etc). Security note: self-edit access is obviously high-risk by design, but basic controls are built in (policy sandboxing, auth, secret management). Tech stack: TypeScript, Electron desktop app + WebSocket gateway, pnpm + Vite/Turbo. Quickstart is literally: pnpm install && pnpm run dev It's very early/WIP, but the self-editing loop already works in basic scenarios and is surprisingly addictive to play with. Would love feedback from folks who have built agent systems or thought about safe self-modification.
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
