Step 1 · Foundations · Foundations · What it is ENPT
Organizer Loop Engineering · Visual Course

What it is

A single place to see what every AI agent is doing and what you should organize or do next.

Read the plain version, or open the technical layer on any section.
1

The big idea

You run multiple AI agents — Claude, Codex, Kimi, Grok — across many Appfy projects. Sessions pile up, emails stay unread, and the real next task hides inside scattered logs, worktrees, and inboxes.

Organizer Loop Engineering turns that noise into one calm command center. It reads the reports already on your machine, builds a data.json of tasks, and serves a local dashboard where you can triage, edit, and decide what to do next.

Think of it like… a personal assistant that collects every sticky note from your desk, sorts them into one inbox, and leaves them ready for you to process. It does not do the work for you; it makes the work visible.

Under the hood

The system is local-first and zero-build. dashboard/collector.py parses Markdown reports under docs/ and sessions/, normalizes them into a stable schema, and writes dashboard/data.json. dashboard/server.py serves static files and exposes an /agents endpoint that reports which agent processes are currently running on this Mac.

2

In one picture

Agents Reports collector.py writes data.json Dashboard triage views
Agents and reports → collector → data.json → dashboard.
3

In the code

The schema is simple: every task has an id, title, project, area, status, priority, source, dates, tags, and notes. The collector turns free-form Markdown into this shape.

dashboard/collector.py 
# A task is a frozen dataclass with one canonical shape
class Task:
    id: str
    title: str
    project: str
    area: str
    status: str
    source: str
    priority: int
    created_at: str
    due_date: str | None
    tags: list[str]
    notes: str

Access it yourself

Open dashboard/collector.py in any editor, or run:

open ~/Documents/Projects/appfy/organizer-loop-engineering/dashboard/collector.py
4

Try it

Generate the data file now. It costs nothing and takes about one second.

cd ~/Documents/Projects/appfy/organizer-loop-engineering
python3 dashboard/collector.py
You are the operator, not the agent. The dashboard only shows what is already on your machine. In the next lesson you will start the server and open the UI. If a task feels missing, check docs/INVENTORY.md first.