AI Coding Agent Tutorials
Let a coding agent validate its own changes against real dependencies, or test agent-written pull requests before you merge them.
| Tutorial | Description |
|---|---|
| Cursor | Use Cursor's agent and the Signadot MCP server to route cluster traffic to code on your machine and verify changes in a closed loop. |
| Claude Code | Change a service locally with Claude Code and validate it end to end through the cluster with routing keys. |
| Codex | Let Codex create a sandbox, run code locally and iterate on its own until a test passes. |
| Devin | Give Devin the Signadot CLI and MCP server so it runs end-to-end tests in your cluster before it reports back. |
| Modal Sandboxes | Run Claude Code in a Modal Sandbox and test its changes end to end against your cluster, with only an API key. |
| Jira Coding Agent | Take a Jira work item to a pull request tested in a sandbox, with Bitbucket Pipelines and Playwright. |
| CodeRabbit | Run integration tests on every pull request, then have CodeRabbit explain and fix the failing tests. |