Ego Lite: Isolated Browser Spaces for AI Agents

Problem and target customer

Why this exists

Customer problem

Browser automation tools force agents into a separate browser that loses login state, or they share the user's active tabs and create contention between human and agent workflows.

Who pays

Developers and operations teams using coding agents like Codex, Claude Code, or Cursor who need authenticated browser automation with parallel task isolation on macOS.

Business model

Free macOS runtime for individual developers; paid team plans for concurrent Spaces, managed browser-state storage, domain approvals, audit history, and enterprise administration.

Editorial note

Ego Lite removes a specific collision in browser automation: a clean browser has no authenticated state, while an agent that takes over the user's active tabs interrupts the developer. Isolated Spaces let an agent work with approved sessions while the developer keeps using the main browser.

The paid product is built around team capacity and control. Individual developers can use the local runtime, while teams pay for more concurrent Spaces, managed browser-state storage, shared domain approvals, audit history, and administrative controls. The product should treat session data as sensitive infrastructure, not as a generic storage upsell.

The MVP stays on macOS, uses a domain approval list as its first guardrail, and proves parallel authenticated workflows. Cross-platform support and a full enterprise policy engine should follow only after the session boundary is documented and tested.

Choose your next step

What do you need next?

Evaluate the operating tradeoffs quickly, or inspect how to build the MVP.

Evaluate this ideaSee the cost and deployment preview

Compare the three operating paths at the default workload before opening the full calculator.

See cost preview Build this ideaInspect the build blueprint

Review MVP scope, workflow, states, data rules, architecture, and acceptance criteria.

Open blueprint

Evaluation preview

What would it take to run?

Directional infrastructure estimates for the current 1,000 browser tasks assumption. Open the full calculator when you are ready to change it.

Open full cost and deployment

ManagedSelected

$5-$37/ month

Managed app, database, and worker capacity for the beta workload.

Lowest operating effort

Lean self-hosted

$5-$27/ month

A lower fixed-cost deployment with more setup and maintenance.

Lowest baseline cost

More control

$20-$92/ month

More control over service placement and operational capacity.

Most separation and control

Build blueprint

Build the first paid use case

Product goal

Who it serves and what it must change

Target user: Developers and operations teams using coding agents on macOS who need authenticated browser automation without tab contention.
Problem: Browser automation tools lose login state in separate browsers or compete with the user for active tabs, making parallel agent workflows impractical.
Measurable outcome: An agent can complete an authenticated web task in an isolated Space while the user browses normally, with page snapshots captured for every navigation step.

MVP scope

What ships now and what waits

Included

macOS browser runtime with shared authenticated profile state
Per-agent isolated Spaces with independent tab contexts
Domain-level approval list for Space access
Page snapshot capture and JavaScript browser tool execution
Agent skill integration for Claude Code or Cursor

Excluded

Windows and Linux support
Enterprise SSO and audit logging
Cloud-hosted browser runtime
Automated cookie redaction

UX and user flow

Screens, actions, and states

Browser home

Shows user's active tabs and a sidebar listing agent Spaces.

Create SpaceOpen existing SpaceBrowse normally

Space setup

Configure a new agent Space with approved domains and inherited login sessions.

Select domainsAssign agentStart Space

Space workspace

Displays agent navigation, snapshots, and task status in an isolated context.

View snapshotInspect task logPause or close Space

Snapshot viewer

Renders a captured page snapshot with agent annotations for review.

Navigate snapshot historyExport snapshotApprove next step

Primary flow

Install browser and agent skill
Create isolated Space with domain approvals
Agent executes browser task in Space
User inspects snapshots and results
Close or archive Space

Loading, empty, and error states

Space created, awaiting agent connection
Agent connected, task running
Task paused for user approval
Task completed, Space archived

Core entity outline

Entities and business rules

BrowserProfile

Represents the user's authenticated browser state shared with Spaces.

Fields: id, user_id, cookies, sessions, created_at
Relations: 1 profile to many Spaces

Space

An isolated agent workspace with its own tab context and domain approvals.

Fields: id, profile_id, agent_id, approved_domains, status
Relations: Belongs to BrowserProfile, Has many BrowserTasks

BrowserTask

A unit of agent work within a Space, such as navigating or scraping a page.

Fields: id, space_id, instruction, status, started_at, completed_at
Relations: Belongs to Space, Has many PageSnapshots

PageSnapshot

A captured render of a page at a point in time for agent or user inspection.

Fields: id, task_id, url, html_hash, captured_at
Relations: Belongs to BrowserTask

AgentSession

Tracks the coding agent connection and skill version active in a Space.

Fields: id, space_id, agent_type, skill_version, connected_at
Relations: Belongs to Space

Business rules

A Space may only access domains on its approved list.
Only one agent session may be active per Space at a time.
Page snapshots are captured on every navigation step within a task.
Closing a Space does not affect the user's active browser tabs.
Authenticated state is inherited from the browser profile only for approved domains.

Architecture and data flow

Components, integrations, and controls

Browser runtime

Renders user tabs and isolated agent Spaces, manages shared profile state.

Space manager

Creates, pauses, and archives Spaces; enforces domain approval rules.

Snapshot service

Captures and stores page snapshots for each task navigation step.

Agent skill bridge

Exposes JavaScript browser tools to coding agents via the ego-browser skill package.

Integrations

Claude Code via ego-browser skill
Cursor via ego-browser skill
Codex via ego-browser skill
Make webhook for task-completion notifications

Data flow

User creates Space and selects approved domains
Space manager inherits authenticated state for approved domains from BrowserProfile
Agent skill bridge receives task instructions and drives the Space browser context
Snapshot service captures each navigation step and stores it locally
Task completion triggers optional Make webhook to external channels

Failure handling

If an agent navigates to an unapproved domain, the Space blocks the request and logs a warning
If a page structure changes and snapshot capture fails, the task is marked degraded and the user is notified
If the agent skill connection drops, the Space pauses and preserves the current task state

Security

Domain-level approval list restricts agent access to authenticated sessions
Authenticated state is scoped per Space and never exposed to unapproved domains
All snapshots and task logs are stored locally on the user's machine

Rate limits

Maximum 5 concurrent agent Spaces per browser profile in MVP
Snapshot capture limited to 1 per navigation step to avoid storage bloat

Deliverables and acceptance

Definition of done for the MVP

Deliverables

macOS browser app with Space creation and domain approval UI
ego-browser skill package installable in Claude Code or Cursor
Local SQLite store for profiles, Spaces, tasks, and snapshots
Snapshot viewer with navigation history

Acceptance criteria

User can create a Space, approve a domain, and an agent can navigate an authenticated page on that domain using inherited login state.
While the agent task runs, the user can browse normally in separate tabs with no tab contention.
Every agent navigation step produces a viewable page snapshot in the Space workspace.
An agent attempting to access a domain not on the approval list is blocked with a logged warning.
Closing a Space does not close or disrupt any of the user's active browser tabs.

Recommended stack

Enough technology for the first version

Desktop app

Electron or native macOS browser shell

Provides the browser runtime that exposes user profile state and renders isolated agent Spaces.

Agent skill

ego-browser skill package

Defines the code-oriented browser functions agents call instead of repeated CLI invocations.

State storage

SQLite local store

Persists browser profiles, Space metadata, task logs, and page snapshots locally on the user's machine.

Automation glue

Make

Optional webhook bridge to route task-completion notifications from agent Spaces to external team channels.

Why this is sufficient

The core product is a local macOS browser runtime with an agent skill layer; no cloud backend is required for the MVP. SQLite keeps sensitive authenticated state on-device. Make is optional glue for notifications, not a core dependency.

Not required for the MVP

Cloud-hosted browser infrastructureLLM inference serviceTeam billing and subscription managementCross-platform packaging

Next stepTurn the blueprint into an execution plan

Copy the build prompt, model the operating cost, and choose where to deploy.

Build and ship

Execution

Build, price, and deploy the MVP

Once the blueprint is clear, use the prompt, cost model, and deployment options to start building.

Build prompt

Copy this into a builder

lovable

Build prompt

Your build prompt is ready

Open the prompt pack whenever you are ready to take this blueprint into your builder.

Based on the blueprintReady for your builderFollow-up steps included

Cost calculator

Model the MVP operating cost

A technical run-cost estimate for the first production version. Team, acquisition, payment fees, and business COGS are excluded.

Estimated monthly total$5-$37

$2.5-$19 per 500 browser tasks

Browser tasks / month1,000 browser tasks

Selected pathEasiest

Pricing checkedAug 2, 2026

Cost breakdown

$5-$37 per month

Low and high values allow for usage variance and plan headroom.

Managed application stack

Managed app, database, and worker capacity for the beta workload.

2K browser tasks included, then $5-$20 per 2K browser tasks

$5-$25

Browser runtime

Desktop runtime for isolated browser Spaces and controlled page interaction.

Monthly allowance from this idea's operating profile

$0-$12

Included

Application hosting
Database and background processing
Required product modules

Not included

Labor
Marketing
Taxes and legal review
Enterprise compliance work

Pricing basis

The estimate combines the selected infrastructure path, required operating modules, selected optional modules, and usage above included monthly allowances. Taxes and regional uplifts are excluded.

Railway pricing Chromium project

Deployment options

Pick the operational tradeoff

Choose based on operating preference, not only the headline price.

EasiestRecommended

Railway

Run the web app, database, and background jobs with a small operational footprint.

$5-$25/month plus usage modules

Good fit

Managed beta
Fast deployment
Background jobs

Limitation

Service usage and background workloads can exceed the base plan.

Deploy on Railway

Cheapest

Vultr

Run the app, worker, and database on one small Vultr VPS with Docker Compose and explicit backups.

Varies by selected stack

Good fit

Hands-on operator
Small pilot
Low fixed cost

Limitation

You manage more setup, backups, monitoring, and upgrades.

Deploy on Vultr View the lean VPS guide

More control

DigitalOcean

Separate application, worker, data, storage, and backup responsibilities as the workload grows.

$10-$60/month plus usage modules

Good fit

Dedicated capacity
Custom services
Operator control

Limitation

Security updates, backups, observability, and scaling remain your responsibility.

Explore DigitalOcean View the separated cloud guide

Start building

Prototype, automate, then deploy

Recommended route

01Vibe-coding builderSpace management UI

Build the desktop interface for creating agent Spaces, selecting approved domains, viewing snapshots, and inspecting task status alongside the user's normal browser tabs.

Open Lovable 02AutomationTask notification bridge

Optionally connect Space task-completion webhooks to Slack or email so teams know when an agent finishes a browser task.

Open Make 03DeploySnapshot and task storage

Deploy a managed SQLite-compatible persistence layer for teams that need shared Space logs and snapshot history beyond a single machine.

Open Railway

Ego Lite: Isolated Browser Spaces for AI Agents

Is this worth validating?

Build this if

Avoid this if

Validate first

Why this exists

Customer problem

Who pays

Business model

What do you need next?

What would it take to run?

Build the first paid use case

Who it serves and what it must change

What ships now and what waits

Included

Excluded

Screens, actions, and states

Primary flow

Loading, empty, and error states

Entities and business rules

Business rules

Components, integrations, and controls

Integrations

Data flow

Failure handling

Security

Rate limits

Definition of done for the MVP

Deliverables

Acceptance criteria

Enough technology for the first version

Electron or native macOS browser shell

ego-browser skill package

SQLite local store

Make

Why this is sufficient

Not required for the MVP

Build, price, and deploy the MVP

Copy this into a builder

Your build prompt is ready

Model the MVP operating cost

Usage assumptions

$5-$37 per month

Included

Not included

Pricing basis

Pick the operational tradeoff

Railway

Good fit

Limitation

Vultr

Good fit

Limitation

DigitalOcean

Good fit

Limitation

Prototype, automate, then deploy

Related ideas

BookToSkill: Turn Technical Documents into Agent Skills

OpenCodeReview: Deterministic AI Code Review for CI

Buzz: Human-Agent Workspace for Auditable Projects

Podcast Voice Cloning & Episode Repurposing Suite

AI Headshot Studio for Professionals & Teams

EnergyMatch — AI Daily Planner for Solopreneurs