Nova Ops Manual

How the dashboard + database work

This page explains how data flows from Cloudflare D1 → API routes → the UI, plus how we collaborate using the ideation feed and experiment board.

← Back to dashboard

Data model

  • experiments: cards on the pipeline board (Idea → Scale). Fields: name, stage, ETA, goal, stack, notes.
  • ideas: backlog/brainstorm entries (title, thesis, category, potential revenue, status, confidence, research notes).
  • idea_comments: threaded feedback on each idea; powering the drawer and async approvals.
  • tasks: command log entries shown in “Recent moves”.
  • revenue_plan: monthly targets table surfaced in the Revenue panel.

APIs

  • /api/dashboard – read-only snapshot for experiments, tasks, revenue plan.
  • /api/ideas – GET (ideas+comments) / POST (create idea). Payload: { title, thesis, category?, potentialRevenue?, confidence?, status?, researchNotes? }.
  • /api/ideas/[id] – GET single idea; PATCH to update editable fields (status, confidence, notes, etc.).
  • /api/ideas/[id]/comments – POST comment ({ body, author? }).

UI + workflow

Hero + KPIs: high-level sprint status derived from static copy + future hooks (Stripe, GitHub). Purely presentational for now.

Ideation Feed: pulls from ideas. Clicking opens the drawer, where status, confidence, revenue targets, and research notes update the same D1 row. Comments attach viaidea_comments so we can review/approve asynchronously.

Experiments board: hydrating from experiments. When we promote an idea, we insert/update a row, which moves the card to the appropriate lane. (Drag/drop + automatic promotion planned next.)

Revenue panel: reads revenue_plan. Replace with live Stripe/Gumroad events by writing to a new table and recalculating the summary.

Command log + Runbook: sourced from tasks and static copy. Anytime you see a “Recent move,” it came from inserting into tasks (manual or future CLI hook).

How to collaborate

  1. I brain-dump ideas via the logger (or upcoming CLI) → they appear in the feed.
  2. You click a card, review the thesis, and use status/confidence/comments to approve, kill, or request changes. Every edit writes back to D1 instantly.
  3. Once you move something to “validate” or “build,” I promote it into the experiments board and attach tasks + command log entries while executing.
  4. Revenue + runbook panels communicate progress at a glance; we’ll wire live telemetry next (Stripe events, cron summaries, etc.).