The current integration focuses on monitoring and management: listing agents, checking container status, exporting usage data, and managing schedules. To launch a specific Phantom with custom inputs, configure it in PhantomBuster's dashboard with a schedule or trigger. The agent excels at oversight, helping you understand what is running, what succeeded, and what failed.

The agent calls PhantomBuster's Fetch Running Containers endpoint, which returns all active containers across your organization. Each container shows which agent it belongs to, when it started, and its current state. Your team gets real-time visibility into active automations without checking the dashboard.

PhantomBuster's agent usage export produces a CSV with run counts, execution durations, success and failure rates, and per-agent breakdowns. Container usage exports add individual run timestamps, exit statuses, and resource consumption. Reports cover 1 to 180 days of history. The download link expires after 30 days.

Yes. The agent can call PhantomBuster's Unschedule All endpoint to disable every scheduled Phantom launch organization-wide. This is a safety mechanism for when you are approaching rate limits, need to pause operations, or want to stop all automated runs immediately. Individual Phantom schedules can be reconfigured in the dashboard.

No. Tars interacts with PhantomBuster's management API, not the scraped data itself. Scraped leads and profiles remain in your PhantomBuster storage or connected CRM. The agent helps you monitor and manage the automation infrastructure, but it does not access or store the actual output data from Phantom runs.

Agents are the automation configurations (your Phantoms). Containers are individual execution instances. When an agent runs, it creates a container that processes the task and terminates. Branches are script versions used in development. The AI agent can list all three, helping your team understand what is configured, what ran, and what code is deployed.

Yes. The Fetch Resources endpoint returns your organization's current quotas including phantom slots, execution minutes, and storage. The agent can check these before advising on whether to launch additional Phantoms. This prevents unexpected limit breaches that would pause your entire automation pipeline.

PhantomBuster includes a built-in AI text generation module. The agent can send a prompt to this module and receive text completions, useful for generating personalized outreach messages for scraped leads. You specify the model, temperature, and max tokens. The generated text can be used to craft LinkedIn messages or email copy at scale.