ZENITH: I Was Too Lazy to Manage My AI Agents, So I Built an AI to Manage Them
This is a submission for the Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady What I Built I have a terrible habit of starting interesting projects. I also have a less interesting habit of coming back to them and asking myself the same three questions: What was I doing? Why did we make that decision? And when did "all tests pass" stop being true? The answers were scattered across…
I discovered a recurring problem: I would start interesting projects, only to forget what I was doing, why decisions were made, and if the tests still held true. To address this, I created Zenith, an open-source, local-first personal intelligence and agent-orchestration platform. Zenith's memory is not just the language model's memory, but the system's memory.
It tracks projects, decisions, unfinished tasks, evidence, source freshness, and execution history. Zenith also provides a workflow for planning, implementing, and independently checking engineering work. The system uses two main ideas: memory should distinguish between claims and facts, and approvals are necessary to confirm success.
Misa is the assistant that sits on top of Zenith, capable of using a local model like Gemma 3 4B through Ollama for reasoning. The system aims to be lazy-friendly, allowing users to walk away from work without having to reconstruct their entire life. A demonstration using fictional projects, Harbor and Lighthouse, shows how Zenith remembers past work, coordinates coding agents, verifies their claims, and requires approval before proceeding.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
