Automation Platform overview

Warp Factories and standalone cloud agents run on the Automation Platform. It provides the execution, configuration, and APIs they share.

For a team workflow that takes issues through planning, implementation, and review, start with the Warp Factories quickstart in the Warp Factories web app. For a single run in Warp, use the cloud agents quickstart. See transitioning from Oz if you already use the legacy web app, CLI, or SDKs.

How a run works

A run starts with a request or event and ends with a result you can inspect:

1. A trigger fires: a schedule, an integration event like a Slack mention or a CI failure, an API call, or a manual start.
2. Warp creates a task, the tracked record of the run. The trigger’s context travels with it: the Slack thread, the PR metadata, the CI logs.
3. The agent executes on a host, optionally inside an environment that defines its image, repos, and setup.
4. The task produces outputs: a pull request, a Slack reply, a report, or just a transcript and summary.

Integrations and triggers

Factory automations route events and schedules to factory agents. Connect the tools your team uses, or use webhooks and factory endpoints for custom intake.

Standalone agents also have integrations and scheduled runs. For CLI setup, use oz integration create on the Oz CLI; see the integration setup guide.

Tasks and tracking

Warp tracks every run as a task: its status, transcript, and outputs stay available after the run finishes. Watch or steer a live run with session sharing, browse history in the management UI, or query it from the Oz CLI and the API. Access control decides who can run, view, or intervene in tasks.

To fan work out across parent and child agents, see multi-agent orchestration.

Environments

An environment defines what a run needs: a Docker image with your toolchain, the repositories to clone, and setup commands. Automated runs (integrations, schedules, API calls) use an environment so every run starts from the same setup; interactive local runs use your machine as-is and don’t need one. Define one environment per codebase and reuse it across triggers.

Hosts

A host is where the agent executes. By default runs execute on Warp-hosted infrastructure, with nothing to set up. On Enterprise plans, self-hosted runners keep code and execution inside your own network while Warp still tracks the runs.

The CLI

The Oz CLI starts and manages runs where there’s no UI: CI jobs, scripts, and remote servers. Start a run with oz agent run, and it reports progress to Warp like any other task, so work that starts on a CI runner shows up alongside everything else your team runs. For interactive sessions, use agents in the Warp app.

API and SDKs

Use the Warp Platform API and SDKs to send work from your own services. Factory endpoints dispatch by factory UID; agent and run endpoints start standalone runs and inspect, continue, or cancel runs.

The official warp-platform-sdk Python package and @warp-dot-dev/warp-platform-sdk TypeScript package provide typed requests, retries, and error handling.

Secrets

Agents often need credentials for APIs, cloud providers, databases, and MCP servers. Store them as secrets, and Warp injects them at runtime without exposing the values in logs or the UI. Secrets can be scoped to the whole team or to one person.

Shared configuration

For factories, declare shared skills, MCP servers, and execution defaults in the factory definition. Standalone workflows also use MCP servers, rules, saved prompts, and environment variables; follow each reference for its scope and setup.

Warp Factories

Use Scorers, benchmarks, and Self-improvement to evaluate factory work and propose changes to its configuration. See measure and improve for metric definitions and limitations.

Related pages

• Warp Factories quickstart - Create a factory and send its first work item.
• Architecture - reference diagrams of the stack, the run lifecycle, and data boundaries.
• Cloud agents - what cloud agents are, how they get triggered, and how to run them with or without the Warp app.
• Cloud agents quickstart - run your first cloud agent in about ten minutes.
• Environments - define the toolchain and repos a run executes against.
• Warp Platform API - drive the platform programmatically.