How to use Unknown to refine Azure DevOps work items
An engineer used an agent-built Azure DevOps CLI to combine cards, meeting transcripts, requirements, and repository context before approving batch updates.
news.ycombinator.comPrompt to try
Editorial adaptation, not the source author’s exact prompt. Replace bracketed inputs and review tool permissions first.
Create a no-op-by-default CLI for [Azure DevOps project] that exports cards as JSON, Markdown, and CSV. Use [meeting transcripts], [requirements], and [repository context] to propose description and acceptance-criteria refinements. Print all updates and require approval before writing.
Reach people exploring AI agents
See placements →What happened
The author reports a CLI that exports Azure DevOps cards and pushes updates, with contextual refinements discussed in a long agent conversation.
Reported by gchamonlive on Hacker NewsHow the agent did it
- Created a CLI to pull Azure DevOps cards in JSON, Markdown, and CSV.
- Loaded meeting transcripts and requirements alongside repository context.
- Discussed refinements, previewed changes, and approved board updates.
Where a person stepped in
The author reviewed the printed no-op preview and approved changes before the CLI wrote to the canonical board.
A person approved each step before the agent continued.
A DevOps engineer described using agents to create a small CLI for planning work. The tool pulled their Azure DevOps cards, exported them as JSON, Markdown, and CSV, and could push updates after refinement. The author used this utility to work through descriptions and acceptance criteria without implementing the tooling manually.
They loaded meeting transcripts and other written requirements into the agent's context, then compared that material with the current state of repositories. A long conversation with the agent explored the cards and identified the refinements needed. Once the changes were ready, the author pushed them together rather than editing each item individually.
The author required their CLIs to be no-op by default. Each first printed the changes it would make, and only committed to the canonical board after the author decided the changes made sense and approved them. This planning workflow sat alongside a broader practice of attentive output review, follow-up refactoring, code-quality checks, and manual testing. The source reports a working utility and a review process, but does not quantify time saved by this particular CLI.
Who it's for: DevOps engineers refining a backlog against meeting notes and the current repository state.
Prompt
Suggested prompt (written by Agents In Use, not the author's)
Source: Hacker News · gchamonlive
Frequently asked questions
What formats did the CLI export?
The author says the CLI pulled Azure DevOps cards and loaded them as JSON, Markdown, and CSV before pushing updates after refinement.
How were unintended board changes prevented?
The author made all their CLIs no-op by default: tools first printed proposed actions, and the author approved sensible changes before committing them to the canonical board.
More details
About this account
The agent built the CLI and helped explore work items and refine descriptions and acceptance criteria through conversation.
Platform: Hacker News. Attribution: https://news.ycombinator.com/item?id=48347996. Prompt: Suggested (written by Agents In Use, not the author's). Sourced by the Agents In Use pipeline.
Credit: gchamonlive on Hacker News
Source published 31 May 2026
Source checked 14 Oct 2026
Tools and topics
AI agent · Azure DevOps · Custom CLI
Azure DevOps · Work items
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.