Getting started

Quickstart

Get up and running with Inspector in minutes. Follow the steps below to set up your CLI and connect it to your Workspace.

⚠

Local executor required

To fully use Inspector and unlock features related to coding and automated task execution, you need to install Codex CLI or Claude Code.

Local executor →

Step 1: Setup CLI

1

Download and install Inspector CLI

bash
curl -fsSL https://cli.inspectorai.pro/install.sh | bash
2

Open your terminal in any project directory and start Inspector CLI:

cd /path/to/your/project
inspector
3

Log in using your Inspector account credentials.

4

Bind project directory you want to work in:

cd /path/to/your/project
inspector bind
5

In the bind dialog that appears:

  • Select or create a Workspace to work in.
  • Select or create a Project to bind to that Workspace.
6

After a successful bind, a local executor is created on your device. Tasks will be dispatched to it for execution.

Tip

Control your local executor at any time with inspector start and inspector stop.

Step 2: Setup Web

1

Create or sign in to your account at inspectorai.pro.

2

After signing in, you'll land in your main Workspace.

3

Go to the Projects tab, select the desired project, and assign yourself as an Assigned Member.

4

On the Board tab, you can manage the entire development process — assign tasks to projects, to team members, and hand off tasks to your local executor using the Hand off to Agent button.

3. Main workflow

1

Chat with the agent

In the Chat section, discuss your ideas with the AI agent. When ready, it will generate documentation for implementing your task.

2

Confirm the spec

If the documentation matches your requirements, click Confirm — the agent will create tasks based on the specification for each relevant project.

3

Assign tasks

Once tasks are created on the Board, assign them to any team member.

4

Hand off to agent

The assigned user clicks Hand off to Agent to send the task to their local executor for automated implementation.

5

Human review

After the task is complete, the agent moves it to Completed status and the Human Review Kanban column for your final approval.