Skip to content

Home

Jump to bottom
DigitalBrain Wiki edited this page Jul 16, 2026 · 6 revisions

DigitalBrain

Your personal operating system — one living workspace where people and AI agents work together, safely.

DigitalBrain treats every useful thing as a Neuron: Chat, Login, Gmail, Salesforce, Features, Approvals, even the screens you see. Humans use a live app. Agents join through the same brain. Nothing important happens “off to the side” in a tool the app cannot see.

Why this matters (in plain English)

Most products split the world:

Old world DigitalBrain
Chatbot over here, apps over there One workspace for chat, work, and integrations
AI agents with their own tools & memory Agents use the same brain the human sees
“Connect Gmail” = install random plugins Connect once → health, permissions, approvals
Automations that surprise you You approve anything that changes the outside world
New features dropped by eng only The product can grow itself — with your approval

Bottom line: DigitalBrain is not “another chat UI.” It is a shared brain you can trust to grow, connect, and act — with people still in charge.

One sentence

DigitalBrain is a personal OS built as a workspace of Neurons: humans and agents share the same state, the app only shows live truth, and external changes need human approval.

Product architecture (the big picture)

This is product architecture, not a server diagram. Read it as: who acts → what lives in the brain → how the outside world is touched.

flowchart TB
  subgraph People["People"]
    H["You in the app<br/>Chat · Features · Connections · Activity"]
  end

  subgraph Agents["AI agents"]
    A["Claude · Codex · Grok · …<br/>via DigitalBrain MCP"]
  end

  subgraph Brain["One shared workspace — the brain"]
    direction TB
    L["Login / Session"]
    C["Chat"]
    F["Features"]
    K["Connections"]
    E["Effects & Approvals"]
    Y["Activity"]
    L --- C
    C --- F
    F --- K
    K --- E
    E --- Y
  end

  subgraph Outside["Outside world"]
    G["Gmail / Google"]
    S["Salesforce"]
    W["Other apps"]
  end

  H -->|"see & act<br/>live surfaces"| Brain
  A -->|"same Neurons<br/>same Owner"| Brain
  Brain -->|"only after you approve"| Outside
  K -.->|"connect & unlock"| G
  K -.-> S
  K -.-> W
  E -->|"prepared change"| G
  E -->|"prepared change"| S
Loading

How work actually flows

flowchart LR
  subgraph In["1 · Enter"]
    I1["Sign in"]
  end
  subgraph See["2 · See"]
    I2["Live screens<br/>project the brain"]
  end
  subgraph Do["3 · Do"]
    I3["You or an agent<br/>drive Neurons"]
  end
  subgraph Trust["4 · Trust"]
    I4["Risky outside changes<br/>wait for Approve / Reject"]
  end
  subgraph Grow["5 · Grow"]
    I5["New Features enter<br/>only when you approve"]
  end

  In --> See --> Do --> Trust --> Grow
  Grow -.->|"brain is stronger"| See
Loading

The “happy path” of power

flowchart TB
  Start(["User job"]) --> ChatN["Chat Neuron<br/>talk / plan / drive work"]
  Start --> Conn["Connector Neuron<br/>e.g. Google healthy"]
  Conn --> Cap["Capabilities unlock<br/>read mail, update CRM, …"]
  ChatN --> Feat["Feature Neuron<br/>repeatable skill"]
  Cap --> Feat
  Feat --> Eff["Effect Neuron<br/>prepared outside change"]
  Eff --> Human{"Human decides"}
  Human -->|Approve| World["Outside system updates"]
  Human -->|Reject| Stop["Nothing external changes"]
  World --> Act["Activity shows the outcome"]
  Stop --> Act
Loading

What the product can do (power map)

Power What a PM should picture Where it lives
Talk & work in one place Chat is not a toy — it drives real work the rest of the app can see Chat surface · Agents via MCP
Humans + agents together An agent acts; your screen updates live. Same workspace, no secret side channel The Shared Workspace
Connect real systems Google, Salesforce, … as healthy connections, not plugin shopping Capabilities & Connections
Stay in control Sends, CRM updates, and similar changes are approval moments Effects & Trust
Grow the product itself New “Features” (skills) can be proposed and installed — only after human decision Building on DigitalBrain
See what happened Activity is the honest history of operations and approvals Destinations in the app
Say no to chaos We deliberately refuse agent marketplaces, dual brains, and silent auto-mutate What We Delete

Everything above is still the same idea: Everything is a Neuron.

A day in the life (story)

  1. You sign in — you enter your workspace (not a generic website session only).
  2. You open Chat — you are looking at live brain state, not a private offline notepad.
  3. You connect Google — a Connection becomes healthy; useful abilities unlock.
  4. An agent helps — via DigitalBrain MCP it works in the same Chat / Features world; your UI lights up.
  5. Something needs to change Salesforce — the product prepares an Effect; you Approve or Reject.
  6. Activity shows what ran, what waited, and what finished.
  7. A new Feature is proposed to make a workflow repeatable — you decide whether it becomes part of the brain.

That loop is the product. Details live on the linked pages.

Hard rules (do not dilute)

Short version for decisions and PRDs:

  1. Everything is a Neuron — not a metaphor only; product nouns map to Neurons.
  2. One workspace — no separate “UI brain” vs “agent brain.”
  3. The app projects live state — it does not secretly own a second model of truth.
  4. Agents are clients — not a tool store and not “paste any MCP URL.”
  5. Self-evolution is human-approved — growth goes through a journaled rail.
  6. Outside changes need Effects + people — safety is product, not a compliance afterthought.

How to read this wiki

If you are a product manager or founder (start here)

  1. This page (you are here) — vision, power, diagrams
  2. For Product Managers — jobs, value, FAQs without jargon
  3. Everything is a Neuron — the core idea
  4. Demo Path / Proof — what to show in a meeting
  5. What We Delete — what we refuse so the product stays clear
  6. Glossary — plain-language dictionary

If you are designing or running demos

If you need honesty about “what’s real today”

Full map

Page What it answers
For Product Managers Value, jobs-to-be-done, how to talk about DigitalBrain
Everything is a Neuron Why this is the product, not a slogan
The Shared Workspace One state for humans + agents
Flutter as Live Surface Why the app is a projector
Agents via MCP How agents join without a second brain
Capabilities & Connections Connect → unlock abilities
Effects & Trust Approvals as first-class product
Building on DigitalBrain How the product grows
What We Delete Explicit non-products
Demo Path / Proof Three demos in plain language
Surface Parity Demos Demo runbook
Glossary Shared language
Repo Reality What’s real vs vision
FAQ Sharp questions
Roadmap Shape Direction themes

Who this wiki is for

Product managers, founders, designers, engineers, and people who drive agents.
It is product shaping — shared language and power — not an API reference and not a rewrite plan.

In one line: One living brain of Neurons; people and agents share it; the app shows the truth; you approve what touches the world.

DigitalBrain

Clone this wiki locally