AI Services Market › Use cases › Software & DevOps
Software & DevOpsDone with Codex 2026-10-30
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.
- 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.
Do this with a skill.
These AI Services Market playbooks do this exact job.