Talos documentation
Your first Android automation task.
Install the host application, connect Android and give the agent a small, verifiable goal. Setup time depends on your ADB connection and model provider.
Before you start
- A Windows or Linux host with an available Talos build that matches its architecture.
- Android platform tools:
adbmust be available to the account running Talos. - An Android emulator or an unlocked phone with USB debugging enabled.
- Your own Anthropic API key for Claude, or an HTTP endpoint implementing the Talos custom-provider protocol.
- A licence or trial key when prompted by the installed build. Licensed installations need access to the licence server.
FFmpeg is optional and used for live screen streaming. Appium is not required. AI provider usage is separate from the Talos licence.
1. Install and activate Talos
Choose an available installer for your host. Windows builds run as a tray application; the Linux .deb package installs a systemd service. Use the instructions for the published package and verify its architecture.
Open the dashboard, normally at http://localhost:7700, and paste your licence key if activation is requested. Need a key? Request a free licence trial and save the key shown on screen.
2. Connect an Android device
Start your emulator, or enable USB debugging on your phone and connect it to the host. Accept the authorization prompt on the phone. Check the connection from a terminal:
adb devicesThe device must appear with the status device, not unauthorized or offline. Open Devices in Talos and check the screen preview. Keep the Android screen unlocked.
An emulator that requires an explicit ADB connection must be connected using its own documented address first. Read the compatibility guidance.
3. Choose your AI provider
Claude
In Settings, select Claude, enter your Anthropic API key, choose an available model and save. The agent sends its goal, UI observations and relevant screenshots to the provider. The optional milestone generator also uses Claude and runs only when requested.
Custom HTTP endpoint
Select the custom provider and enter its URL. Your service must accept Talos’s task context, tool definitions, history, observation and optional base64 screenshot, then return a tool decision. This is a Talos-specific protocol, not a drop-in setting for any chat API.
A minimal response shape is:
{
"tool": "tap",
"args": { "element": 8 },
"reasoning": "Open the visible form",
"input_tokens": 0,
"output_tokens": 0
}The element number must refer to the current observation. Custom-provider charges are not calculated by Talos; monitor them at your endpoint.
4. Create a small task with a visible result
Open Tasks, choose New task, name it and select a connected device without an assigned task. Enter a goal and choose one-time mode. Add optional milestones manually or ask Claude to suggest them.
Your goal
“In Contacts, add Alex Morgan with the number 5550100. Confirm the saved contact appears in the list.”
- 01Open the contact form
Success: name and phone fields are visible.
- 02Enter the contact details
Success: the fields show Alex Morgan and 5550100.
- 03Save and check the result
Success: Alex Morgan appears in Contacts.
Illustrative task definition, not a recorded run. The final model call can exceed the budget threshold.
Start with the default limits or adjust iterations, time and estimated cost for the run. The budget is checked after a model response, so the final call may exceed your threshold.
Product screenshots use the Polish UI: Devices is “Urządzenia”, Tasks is “Zadania”, New task is “Nowe zadanie”, and Settings is “Ustawienia”.
5. Start the task and review its timeline
New tasks are created stopped. Start the task from its card. Open the device preview to follow the screen, or use Stop to interrupt the run.
Open run history to review observations, the model’s reported explanations, tools and arguments, results, token usage and available screenshots. Talos retains the last 10 runs per task. Verify the final screen before relying on the outcome.
Repeat a workflow in continuous mode
Choose continuous mode and set a delay from 10 to 86,400 seconds. After a successful cycle, Talos stores a summary and schedules the next cycle. That summary helps the agent continue from context.
Limits apply to each cycle and costs add up. An error, cancellation or reached limit stops the task. Stop also cancels a scheduled next cycle. Stop and finish a running task before editing its definition.
Memory is a model-generated summary, not an exact event ledger. See a recurring status-check example.
If the first run does not work
- No device in Talos
- Check
adb devicesunder the account running Talos. Authorize USB debugging and confirm the cable or emulator connection. The Linux service may run under a different account from your terminal. - The agent cannot reach a screen
- Unlock the device, check app permissions and inspect the timeline. Try a shorter goal with one visible success criterion.
- A task stops at a limit
- Review the recorded result before raising limits. Look for a repeated screen, an unavailable option or ambiguous instructions.
- The preview does not stream
- Check FFmpeg and the live-stream setting. Use the regular device preview to diagnose the connection.
- A licence check fails
- Check your key, activation slots and connectivity to the licence server. The current client does not verify offline licence tokens.
For help, contact the Talos developer with your host platform, Android version and a description of the failed step. Remove private screen content and keys from anything you share.