Guides for managing a delivery backlog with AI agents.
Plain-language guides for people who manage a software backlog.
-
How Dutopilot builds Dutopilot
Dutopilot's own product is planned, built, reviewed and merged by Dutopilot agents from its project board. The setup, a ticket's life, and the real numbers.
-
Can AI agents review pull requests safely?
Yes, as one layer among several. What AI review catches, the risks to plan for, and a safe setup with tests, branch protection and people.
-
How can AI agents work through a Jira backlog?
Yes, if the backlog is a queue of small, clear tickets. How to sort it, the steps an agent follows, and where people stay in charge.
-
Connect Claude to Dutopilot
How to connect Claude, and other MCP assistants, to your Dutopilot projects: the tools they get, what stays read-only, and how your chat history stays yours.
-
How to go from a Jira ticket to a merged pull request with AI agents
The steps an AI agent follows to turn a ticket in your backlog into a reviewed, merged pull request, and where people stay in charge.
-
Running AI coding agents on your own Claude Code, Codex, OpenCode or Oh My Pi subscription
Why many teams bring their own Claude Code, Codex, OpenCode or Oh My Pi subscription or API key to AI delivery, and what it changes for cost and control.
-
Choosing which steps of delivery need a person's approval
A simple way to decide where a person must approve work in AI-assisted delivery, and where agents can move on their own.
Customers: how Dutopilot delivers for its first client projects