-
Notifications
You must be signed in to change notification settings - Fork 0
How it works
Short detail behind the Home story. Still product language — not an engineering manual.
See also: Everything is a Neuron.
| Who | Door | Same brain? |
|---|---|---|
| You | Sign in → app | Yes |
| AI agent | DigitalBrain MCP | Yes |
Different sessions are doors, not second copies of Chat or Gmail.
flowchart LR
App["App<br/>UI Neurons"] --> W["Workspace<br/>Neurons"]
MCP["Agent MCP"] --> W
W --> Screens["Live projections"]
Screens --> App
W --> World["World Neurons<br/>Gmail · Salesforce"]
Screens are not mini-apps with private truth. They are UI Neurons:
| Destination | What it is |
|---|---|
| Login | UI Neuron (and the gate into your workspace) |
| Chat | UI Neuron projecting the Chat work Neuron |
| Features | UI Neuron for skill Neurons you install or author |
| Connections | UI Neuron for World Neurons (Gmail, Salesforce, …) |
| Activity | UI Neuron for what ran, waited, finished |
Buttons and actions (e.g. Approve / Reject) invoke Neurons with a clear target and intent. They are not a separate “click framework” outside the brain.
If an agent changes something, the window should update. That is co-presence.
- Open Connections (UI Neuron).
- Connect Google / Salesforce / ….
- Each system is a World Neuron — health becomes Healthy (or Needs reauth / Disconnected / Misconfigured — be honest).
- Unlocked capabilities appear only when healthy.
- Chat and Features can use those abilities under grants.
- Writes still need approval (next section).
Not: “install an agent package” or “paste a Gmail MCP URL.”
Gmail is a Neuron. Salesforce is a Neuron. Same family as Login and Chat — different job.
Anything that changes the outside world:
- Brain prepares an Effect Neuron (clear summary).
- You Approve or Reject (UI action on that flow).
- Only then does the World Neuron act.
- Activity shows the result.
New Features (skills) use the same spirit: propose → you decide → install.
A Feature is a Neuron with real logic — not a cosmetic toggle.
Because the product exposes UI contracts and connector (world) contracts as the same Neuron model, a Feature can:
- participate in surfaces you already use (Chat, Activity, …)
- call Gmail / Salesforce / … under grants you approved
- run multi-step, complex behavior inside DigitalBrain
- never skip the Effect → Approval rail for outside mutation
flowchart TB
F["Feature Neuron<br/>your skill / logic"]
F --> UI["UI contracts<br/>project · actions"]
F --> G["Gmail Neuron"]
F --> S["Salesforce Neuron"]
F --> E["Effect Neuron"]
E --> You["You Approve / Reject"]
You --> G
You --> S
What this means for you:
- You are not limited to “chat with tools.”
- You can grow the product toward the workflows you want.
- Custom logic still lives in one brain, visible to people and agents, with the same safety rules.
Growth path:
Propose a Feature → You decide → Install → Pause / Resume / Rollback if needed
Authoring richer Features (complex logic on UI + world contracts) is the same product idea, not a second platform.
Agents are clients. They may:
- Drive Chat (same conversation you see)
- Inspect / help with Features
- Propose growth
- Invoke the same Neuron contracts the app uses (through DigitalBrain MCP)
They must not:
- Own a shadow memory the app cannot see
- Silently mutate Gmail / Salesforce
- Become a marketplace of random tools
- Bypass UI or approval rails
flowchart LR
A["1 Connect<br/>World Neurons"] --> B["2 Abilities unlock"]
B --> C["3 Work in UI + Chat + Features"]
C --> D["4 Outside change?<br/>Effect → Approve"]
D --> E["5 Activity trail"]
Optional branch at any time: grow a Feature (including one with complex logic) → same approval spirit → it becomes another Neuron in the graph.
In one line: Same brain; UI Neurons as the live window; World Neurons for Gmail/Salesforce; Features program on those contracts; you approve the outside world.
Next: Demos · Everything is a Neuron · Back: Home