Build a local app for agent output cleanup

They used Codex to create Mianotes, a local place to collect agent-made files and links.

Done with Codexยท Advanced

What they did
The creator spent a year using Codex and OpenClaw on a MacBook and Mac mini, and the files from those agents ended up scattered across the system. They then brought in four Codex agents for more than 30 sessions to build Mianotes, covering the core app, APIs, an MCP server, a dashboard, installers, documentation, and a marketing site. As part of the build, the agents wrote product docs inside Mianotes itself so the workflow could be checked in real use.
How it went
The first public version was ready in two weeks, measured from the first commit to the first macOS package. The app is meant to centralize agent output and convert files and links into Markdown. No usage figures or broader results are given.
Tips
This approach involved a lot of coordinated work, so it is better suited to a bigger project than a one-off script. Before you let an agent move or remove anything, make it stop and ask for approval.

Try it yourself with Codex

I have AI-agent output scattered across [folders/tools]. Help me design and build a local app in [language/framework] that keeps those files and links together, records their source and date, and converts them to Markdown when useful. Begin with a plan and do not touch, move, delete, or submit anything irreversible until I say yes. After approval, build the app, add whatever backend pieces are needed, make a basic dashboard, write clear setup instructions, and include a small test workflow using sample files. Stop and ask again before any payment, deletion, or publishing action.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on reddit.com โ†—

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.