Councils of agents

#267 – November 09, 2025

powerful for senior leaders dealing with complex, nuanced decisions

Councils of agents
9 minutes by James Stanier

Leaders can create councils of AI agents to improve decision-making and accelerate thinking. Using tools like Claude Code, you can build multiple specialized agents that represent different roles like engineers, executives, or domain experts. These agent councils help you test ideas, get diverse perspectives, and iterate on decisions before meeting with real teams. This approach is particularly useful for senior leaders dealing with complex, nuanced decisions across multiple disciplines.

Build Apps Faster with AI-Powered UI
sponsored by Progress Telerik

Tired of repetitive UI code? Now you can build sleek, functional Blazor apps faster — and smarter. With 120+ ready-to-use components and an AI Coding Assistant built right into your IDE, you’ll spend less time tweaking and more time shipping.

Become the consequence
13 minutes by Michael Lopp

The key to senior leadership is becoming "the Consequence". This means consistently reinforcing important company goals and values through regular check-ins with your team. When implementing major changes like improving product quality, you set clear expectations and create ongoing accountability meetings. Rather than micromanaging, you listen to team feedback and adapt your strategy while maintaining focus on core objectives.

How to run 1:1s as engineering manager
10 minutes by Can Duruk

Can gives away a simple, repeatable structure that he uses every single week for his 1:1s. It’s based on the popular People, Product, Process format. He shares the exact script he uses, a template, and some ideas for how to handle each section.

The new calculus of AI-based coding
11 minutes by Joe Magerramov

Agent coding brings new problems. More bugs reach production since teams ship changes much quicker. Traditional testing and deployment systems can't handle the increased velocity. Teams also struggle with communication when making many more decisions per day. Joe suggests the solution isn't just using AI to write code faster, but rebuilding the entire development process around this new speed.

Retrospectives
10 minutes by Lucas F. Costa

Teams often use retrospectives to document problems without actually fixing them. This creates a false sense of progress while issues pile up. Toyota's production system offers a better approach. Workers can immediately stop production when they spot problems, then work together to find root causes and implement permanent fixes. Lucas thinks software teams can apply this too.

Your 2025 QA Playbook: Practical AI Tactics From Engineering Leaders
sponsored by BrowserStack

Want to see how teams are actually getting faster with AI? Join the QA Leadership Summit on November 12 for hands-on examples, AI agents that write tests, catch issues earlier, and cut repetitive work. Leaders from GitHub, Nasdaq, and Cognizant break down what's working in real teams. No jargon, no hype. Save your spot!

industry

security

And the most popular article from the last issue was:

newsletters