Control Center

See the workspace, tell Automous what to do, and stay in control.

Checking AI…
Operating mode Assist AI can analyze and prepare actions, but approval gates remain enabled.
Workspace vision

Dual monitor view

Manual capture in V0.1
1
Monitor 1Procede / Excede
Monitor 1 capture
Procede / Excede workspace Capture or upload a screenshot so the AI can inspect this screen.
AI viewing monitor 1
2
Monitor 2DTNA / OWL
Monitor 2 capture
DTNA / OWL workspace Capture or upload a screenshot to test cross-screen analysis.
AI viewing monitor 2
Automous AI

Command chat

A

I’m ready. Capture one or both monitors, then tell me what you want done.

V0.1 can analyze screenshots and simulate the action plan. It will not control Windows yet.
Live execution

Task plan

Idle
✓
No task running Your next instruction will appear here as a step-by-step execution plan.
!
Review required before final actions Submission, deletion, payment, sending, and other consequential actions should require confirmation.
Coming next

Reusable workflows

Save repeatable desktop tasks such as “Create DTNA claim from RO”, “Audit open repair orders”, or “Copy estimate notes into customer communication.”

01Read source appProcede / Excede
02Validate dataRules + AI
03Fill target appDTNA / OWL
04Human reviewApproval gate
Session log

Activity history

V0.1 stores the current browser session locally. Future builds can persist audited action logs per user and workstation.

AI configuration

Settings

AI backendOpenAI API through Cloudflare Pages Function
Checking…
ModelConfigured server-side with OPENAI_MODEL
gpt-5-mini default
Approval gatesRequire review before consequential actions
Enabled
Desktop controlRequires future Windows companion agent
Not installed
Security

V0.1 guardrails

API keys are never stored in browser code. The Cloudflare Function reads the key from an encrypted environment secret.

  • Same-origin API route
  • Image size reduction before upload
  • No silent desktop control
  • Review-before-submit enabled by default