Coming Next: Building the Mission Control
The first series, the Agent Fleet Series, was about setting up the fleet: an orchestrator, a workforce of specialists, shared memory, and the mistakes along the way.
The second series is about what you do with a fleet once you have one. We build the Mission Control: the dashboard where you see your agents, their tasks, their schedule and their shared memory.
What's different about this series
Almost nothing is typed by hand. We build it with prompts. You will see the prompt, what came back, and what we had to correct. Not a polished end result.
You'll watch the orchestrator delegate. We give the orchestrator one goal. It does not do the work itself. It hands the pieces out:
- Scout researches: what the dashboard should show, and how similar tools do it.
- Dev builds: the dashboard and its pages.
- Scribe writes: once the build is done, Scribe documents what was built.
What we'll build
The dashboard shell first, then a few of its pages, for example an overview, an agents view and a task board. Each page starts as a prompt and ends as something you can open in a browser.
Who it's for
If you followed the Agent Fleet Series, this is the next step. If you didn't, the short Hermes install pre-read gets you to the starting line, and Part 1 explains why a fleet exists at all.
The drafts are in progress. We'll publish the first post here, on X and on the Effibit blog. Follow along, and tell us which page you want to see built first.
New here? Start with the Agent Fleet Series → · Read on Substack