🔧 Setup

Welcome to tappi

Let's configure your AI agent step by step.

Step 1: LLM Provider & API Key

Step 2: Choose a Model

Step 3: Workspace Directory

All file operations are sandboxed to this directory.

Step 4: Browser Profile

Each profile keeps its own logins and cookies. The agent uses this browser to browse the web, fill forms, and interact with sites on your behalf.

🌍 Log in to your accounts

Click the button below to open the browser. Log in to any sites you want the agent to access — Gmail, GitHub, social media, etc. Your sessions will be saved to the profile you just selected.

This is optional — you can always do it later from the sidebar.

Step 6: Permissions

Chat

Connecting...
✕
📎 Drop files here

Browser Profiles

Your Profiles

Each profile has its own browser sessions (cookies, logins) and CDP port.

Loading...

New Profile

Scheduled Jobs

Cron Jobs

Recurring tasks the agent runs automatically.

Loading...

Recent Runs

Loading...

💡 Create jobs via chat: "Schedule a task to check my email every morning at 9 AM"

Job Run

Settings

LLM Provider & Credentials

Model

Workspace

Permissions

(breaks complex tasks into focused subtasks)

Set this to connect to a browser already running on a specific port (e.g. OpenClaw browser). Overrides the profile's default port.

Max time per LLM call. Orphaned threads are killed after this.

Chat responses + compiler/orchestrator in deep research.

Research sub-agents. Lower = faster, more focused.

📁 Choose Workspace Directory

✕
⬆ /