Focus on the what
AI handles the how
Work with remote agents and your team to prepare Runbooks with powerful features designed for modern development workflows.
Plan together, execute with AI
Work with remote agents and your team to prepare the Runbook. Focus on the what while AI handles the how.
Bring your own agents
Use Aviator internal agents or bring your favorite CLI agents, whether it's Claude Code, Gemini, Aider or Codex.
Context aware
Agents read through code and context to understand requirements. Provide global context including coding guidelines.
Secure
Pass Runbooks to AI agents running on your infrastructure. Setup remote MCPs with guardrails.
Preconfigured environment
Every remote environment comes ready with your tools, dependencies, and configurations.
Automation with control
Let AI agents run the entire Runbook autonomously. Or take control by running step by step.
Built for real use cases
From simple features to complex migrations, Runbooks handles the tasks that matter most to your team.
Simple well defined features
Features with small scope, e.g. adding a Slack integration, or defining a new API. Agents may ask clarifying questions.
UI improvements
Share images or mocks to implement new UI or improve existing workflows while maintaining compatibility.
Bug fixes
Give a stack trace or explain the bug. Agents analyze code to root cause and propose fixes.
Improve build times
Using Bazel, Nx or Gradle? Agents can review dependencies and create plans to simplify build processes.
Flaky test resolution
Agents can run tests multiple times to pinpoint issues and propose resolutions for test flakiness.
Code migration
Automate tedious migrations between frameworks, libraries, or language versions with step-by-step Runbooks.
How it works
A simple four-step process that transforms your ideas into production-ready code.
SCAN
Describe the project task. Agents scan through code and documentation, then ask clarifying questions.
PLAN
Agents prepare the Runbook. Invite teammates to collaborate, review steps, and provide feedback.
EXECUTE
AI agents execute steps one-by-one or handle full execution. Agents connect with GitHub to prepare PRs.
VERIFY
Review code carefully, validate changes, and ask agents to write tests. Configure environments to build and run tests.
