Build a local knowledge app for agent outputs with Codex
The author built Mianotes using four Codex agents across 30+ sessions to centralize scattered files agents produced, and the agents documented the product inside it.
Done withCodex
- What they did
- The author built Mianotes after a year of working with Codex and OpenClaw on a MacBook and Mac mini and ending up with agent-generated files scattered around. Using Codex 5.5 Pro, they had four Codex agents work across more than 30 sessions to build the app, its APIs, MCP server, dashboard, installers, docs, and marketing site, while the agents also wrote product documentation inside Mianotes itself to test the workflow.
- How it went
- The first public version was completed in two weeks, from first commit to first macOS package. The post says it centralizes agent output and converts files and links to Markdown, but it does not give usage metrics or broader results yet.
- Worth knowing
- The clearest constraint is effort: four Codex agents and 30-plus sessions were used over a two-week build.
Try it yourself with Codex
I have files that my AI agents produce scattered across [folders or tools]. Build me a simple local app in [language or framework] that collects them in one place, lets me search and tag them, and shows the source and date of each file. You're done when I can run it locally, import a sample of my files, and find one by keyword, and add a README explaining how to use it. Show me the plan before you move or delete any of my existing files.
Want help getting this working? Join the Vibe Coding Academy ↗ Run by Alex Finn, who also runs this site.
Source: Reddit · Undated
Related use cases
Turn tickets into merged pull requests
Pulls tasks from GitHub Issues or Linear, then runs Claude Code or Codex in isolated Kubernetes pods to carry each one through to a PR.
Diagnose a killed process on a server
Codex CLI over SSH found the OOM killer in kernel logs on a 2 GB droplet with no swap and listed fixes by risk without changing anything.
Find bugs with property-testing agent swarms
Agent swarms doing property testing found about 30 bugs in the author's work code, then ~100 bugs across open-source projects like uv, with many fixes merged.
Build a C compiler with a team of parallel agents
Anthropic's engineering write-up on tasking Opus 4.6 agent teams with building a C compiler.
Migrate an Electron app from Vue 2 to Vue 3
Kearney gave Devin repo access to migrate a production Electron app to Vue 3 and update npm packages, committing after every five components.
Auto-generate a searchable help video library for an app
Used Replit Agent to build a /help page that generates tutorial videos via scenes, Playwright recording, TTS narration and ffmpeg stitching.
Get 5 like this every morning
The best things people got an AI agent to do, each with the prompt to try it.