With the latest tools in the mabl cloud MCP, you can now investigate your recent deployments from your AI client of choice. Go from triggering a deployment to digging into specific failures, all without jumping around the mabl UI.
Investigate deployment results
After a merge or deploy, you can go from "did anything break?" to "here's how to fix it" without switching tools or searching for commit hashes. Start with a single prompt, like "What's happening with the deploy I just sent to prod?", and your agent can:
- Poll the in-flight deployment and report live pass/fail statuses as tests complete
- Summarize results when they finish, including failures
- Investigate failed runs and surface evidence, including runtime recovery attempts and links to failed artifacts
- Provide your coding agent with context that can help you fix issues or update tests
Call on the mabl MCP to retrieve the content of specific test run artifacts, such as screenshots and network logs, for more in-depth troubleshooting.
Ask follow-up questions
With new results analysis tools, you can also ask open-ended or follow-up questions about a deployment event, plan run, or test failure to take your analysis a step further. Try prompts like:
"What changed since the last passing run?"
"Walk me through what failed on staging today."