← Cycles
design · c134activestarted Nov 9Auto draft — not yet reviewed

Control Plane Web App

Waiting to be picked uplast activity Nov 10· waiting on Tyler's repo decision and Kamal's availability

Next: Phase 1, the shell mock (sign-in, org, the qdash home with the global filter); Phase 2, the contract written for Kamal after the call.

Why we did this

The company dashboard is stuck inside the program on each laptop, and its sign-in cannot be run by a customer on their own servers.

What happened

The dashboard becomes its own web app, shipped as one package we can run or a company can run itself; both apps are mapped and the backend's 30 gaps are listed.

New in this cycleOne package of software, run by us or by the company itself, that every laptop reports to.
At a glance
3
changes
1
conversation
8
decisions
0
outside helpers

What changed

The two changes that matter most, in plain words. Open any one to see it.

  1. 1

    One package of software, run by us or by the company itself, that every laptop reports to.

    One image, two deployments · came from Tyler Flint, Rob Genova and othersopen →

  2. 2

    We listed everything the two existing apps have, and the 30 things the backend still needs to serve.

    Both apps mapped before anything is proposed · came from Mark Parsonopen →

Also in this cycle
  • A prepared conversation with the backend engineer: what we need, what we can show, what we would accept. In words

How we got here

Each change, traced back to the conversation it came from. A solid mark means the record names the source; a dashed one was matched by the draft and waits for review.

Someone said
TFTyler FlintNov 5matched · review

“bundling in the web project increases the memory usage... for for to just to run a web stuff, like, to run web stuff, and it's already kind of, like, weird.”

market-analyst /calls/1b21a3ce… (local)
TFTyler FlintNov 5cited

“let's strip the data model out, and let's put the new data model in... it would be packaged up into a single Docker file that includes the UI and the Golang app.”

market-analyst /calls/1b21a3ce… (local)
We decided

c134-D1 qdash's shell (the global filter, the when: pill, the nav ruled through c92–c128) inside app.qpoint.io's auth and org frame. — Mark Parson

c134-D2 OIDC plus local accounts for self-hosted, the existing SAML routes kept. No Firebase. — Mark Parson

c134-D3 One org per self-hosted install; SaaS runs one image per customer at first. Multi-org in one instance later. — Mark Parson

c134-D8 The plane exposes a connect endpoint (base URL plus credential exchange), holds mechanic-access grants (c136 D1) and the operator's message inbox (c137 D1). All three are in the API contract. — Mark Parson

So now

One package of software, run by us or by the company itself, that every laptop reports to.

Someone said
TFTyler FlintNov 5matched · review

“bundling in the web project increases the memory usage... for for to just to run a web stuff, like, to run web stuff, and it's already kind of, like, weird.”

market-analyst /calls/1b21a3ce… (local)
We decided

c134-D1 qdash's shell (the global filter, the when: pill, the nav ruled through c92–c128) inside app.qpoint.io's auth and org frame. — Mark Parson

So now

We listed everything the two existing apps have, and the 30 things the backend still needs to serve.

Thanks to

People outside Qpoint first. Every thank-you points at something they said or made.

TF
Tyler Flint
Qpoint team

Took the UI out of the sensor and put the plane on one image

Bundling the web UI undermined the lightweight sensor; the control plane ships as SaaS and as a self-hosted Docker package on Postgres and ClickHouse, Pulse reused with a new data model, no Firebase. A connected qcontrol takes no local configuration. His idea of an OAuth-style connect flow became the connect endpoint in the contract.

“bundling in the web project increases the memory usage... for for to just to run a web stuff, like, to run web stuff, and it's already kind of, like, weird.” internal
RG
Rob Genova
Qpoint team

Kept SaaS from blocking the self-hosted start

Cautioned against rushing a production SaaS for critical security workloads, and ruled that SaaS production-readiness does not block adoption: customers can start self-hosted.

“we don't wanna jump too quickly into, like, 20 customers managing this, like, really critical, like, security stuff... We got, like, 5 guys on the team total.” internal
MP
Mark Parson
Qpoint team

Mapped both apps before proposing anything

Kicked the cycle off with Phase 0: app.qpoint.io and the qdash UI and backend read from the trees, the 30-item API gap list, the asks consolidated, what the layer lacks handed to c135; then the brief for the call with Kamal. Confirmed D1–D8 and ruled that this web app is the layer consumer c135 keeps green.

All images

Every picture from this cycle. Click any one to step through them all.

The pairs

One image, two deploymentsNew in this cycle

Everything captured · 1

The full record 8 decisions · 5 key-change entries · 1,601 words · 1 captures

The record stays as written, for agents and for audit: Key Changes, every decision, the artifacts and the visual record.

Open the record for c134 →

Qpoint Brand Style Guide