Diagnose and roll back bad config during outages
Google SREs use Gemini CLI to pull logs, find a bad config push, generate a reverting changelist, and file postmortem action items via MCP.
Done withGemini
- What they did
- In a simulated outage, an SRE opened Gemini CLI, which pulled incident details, correlated time series and analyzed logs through Google's internal ProdAgent tools. It proposed a restart playbook, and the SRE approved it. The restart failed, so the agent checked the service's source and recent changes, found a bad config push, and drafted a reverting changelist. A custom command then wrote the postmortem and filed action items via MCP.
- How it went
- The agent found the faulty config push in under two minutes, and the SRE-approved revert rolled out and the service recovered. The article gives no other measurements, and the scenario was simulated.
- Worth knowing
- Humans approved each production change, and the agent could only use typed, policy-checked tools. Several tools here are Google-internal, so you'd need your own MCP servers.
Try it yourself with Gemini
An outage is happening in [service name]. Pull the recent logs from [log source or command], find the most recent config change that lines up with the errors, and explain your evidence. Then draft a revert of that change as a patch or pull request, and write a list of postmortem action items. Stop once the revert and action items are ready, and show me both before you push, deploy, or file anything.
Source: cloud.google.com · Undated
Five of these in your inbox every morning
The best things people got an AI agent to do, each with the prompt to try it.
More like this
Race three models to build a Counter-Strike clone
InstantDB had Codex, Opus, and Gemini each try to build a multiplayer shooter.
Edit existing video clips with plain-English instructions
A creator tested Gemini Omni Flash over 36 hours, editing clips in place from typed instructions while preserving lighting and camera angle, with variable quality.
Solve open math problems autonomously
DeepMind's Aletheia agent on Gemini Deep Think was evaluated on 700 open Erdős problems and autonomously solved four.
- Claude
Build a C compiler with a team of parallel agents
Anthropic's engineering write-up on tasking Opus 4.6 agent teams with building a C compiler.
Write and debug code across eight test tasks
A reviewer had Manus build a currency converter, fix a find_max bug and make a landing page, with mostly strong results.
Build and deploy a Game of Life simulation site
In a vendor launch demo, Devin built an interactive Game of Life website, added requested features step by step, and deployed it to Netlify.