Let's connect
Home

AI Automation Dashboard / 2026

ControlDeck

AI Automation Dashboard

Sanitized ControlDeck dashboard preview

Case study

The project at a glance

in progress

Overview

ControlDeck is a local-first command center for monitoring, approving, and safely running personal AI agents, automation scripts, build checks, and workflow tasks.

Problem

Personal automation work can become scattered across notes, scripts, chats, and dashboards, making it hard to know what is safe, active, blocked, or already audited.

Solution

Build a command-center interface with fixed agent definitions, approval gates, demo/local modes, execution logs, and a persistent audit trail.

Outcome

The project has a documented desktop boundary, safety model, and demo proof path. Installer/release proof is intentionally not claimed until the Rust toolchain and Tauri build chain are validated.

Local-first command center for personal AI agents and automation scripts.
Tauri desktop app direction with Rust-enforced approval and safety rules.
Demo/local runner modes, audit events, run history, and approval gates.

Tools used

Stack and proof

TauriReactTypeScriptRustSQLiteNext.js

Public evidence boundary

Desktop direction and proof docs exist locally; release proof is still blocked until Rust/Cargo and Tauri build checks pass.

Role

Product design, frontend, workflow architecture

Case study system

What this project proves

ControlDeck is treated as the flagship product case study: a local-first command center for personal AI agents that makes automation visible, gated, logged, and safe enough to demo.

in progress

Runtime

Tauri

Windows desktop direction with React, TypeScript, Rust, and SQLite

Safety

Allowlist

Fixed Rust manifest owns agents, script keys, and approval rules

Mode

Demo first

Local execution remains gated and disabled by default

Proof

Partial

Desktop direction is documented; release build is blocked by missing Rust toolchain

Process

Build path

  1. 01

    Frame the operating problem

    Personal AI work was spread across prompts, scripts, browser tabs, folders, and terminal commands, which made it hard to audit or safely repeat.

  2. 02

    Design the safety model

    The app blocks arbitrary command input, uses fixed agent definitions, separates demo and local modes, and routes medium-risk actions through approval gates.

  3. 03

    Move toward a real desktop boundary

    The project direction shifted from a web dashboard idea into a Tauri desktop app with Rust commands and app-data SQLite.

  4. 04

    Package it as portfolio proof

    The next presentation layer focuses on public-safe screenshots, a demo video, and clear proof artifacts instead of exposing private local scripts.

Evidence

Source map

What can be shown publicly, what stays local, and what still needs proof before launch.

Public-safe dashboard preview

Public-safe

Sanitized visual asset now shows the command-center concept without exposing private scripts or local data.

Desktop product README

Local-only

Local product description, safety model, demo/local mode rules, and expected release outputs.

Demo proof artifact

Local-only

Local markdown proof records demo mode, no arbitrary command execution, and project-local artifact writing.

Release executable

Pending

Not public yet because the Windows installer and executable proof are still blocked by the missing Rust/Tauri build chain.

Next

What needs to happen before this is final

  • Capture a public-safe dashboard screenshot from seeded demo state.
  • Record a short demo video showing approval and audit behavior.
  • Add the installer/build proof only after the Tauri dependency chain is fully validated.

Visual evidence

Public-safe presentation assets

Sanitized ControlDeck dashboard preview