How I built a PR dashboard that flags risky pull requests
02 Jul 2026
During stand ups we have a PR dashboard which is a list of open PRs, who is assigned and whether it is in a state of approval or not, as well as listing a bunch of draft PRs no one really cares about until they are published for real. So I set claude to work to give me something better.
The tasks:
List the following:
The tasks:
List the following:
- Open PRs
- The repo it lives in
- The author
- The reviewer with review state
- How many tests were added
- How many lines of code were touched
- How many files were touched
- Flag red amber or green large PRs with not many tests added (Threshold: green ≥ 2 tests / 100 product lines)
Claude eventually produced a lightweight front end that communicates to Azure Dev Ops via a Personal Access Token.
The RAG section is just an indicator not to be taken as gospel but it gives the team pause for thought at stand ups for particularly large changes or changes with weak unit test coverage. It rebuilds a unified diff locally for the relevant file types (.cs, .ts, .feature) and counts three things from the added lines: test attributes added in test files ([Test], [Fact], [TestCase], [TestMethod], [Theory]), Gherkin Scenario: lines added in .feature files, and product lines added in non-test .cs/.ts files.
It's not an exact science but it is more useful than the initial list we were using.
Justin Holsgrove
Senior Test Engineer/ QA Manager
He/Him
🤖 Test automator, engineer and QA Manager
✒️ testtechie.co.uk -> testtechie9.wordpress.com
🙆♂️ jholsgrove.github.io
Open To
Meet at MoTaCon 2026
Rosie Sherry
What you say about 'it's not an exact science' is important.
Critics of AI will often point how and where it is wrong, but if it's better than what existed before, then that is progress.
Justin Holsgrove
I will likely make tweaks to it and iterate as needed as I get feedback
Sign in
to comment
SmartBear's vision for using AI agents to support application integrity and software quality
Explore MoT
QALS Summer 2026: a leadership summit to move beyond AI testing pilots and build production-ready, AI-first QA organizations - powered by the BrowserStack AI Test Platform and 25+ connected AI agents
Unleash the power of generative AI to boost your software testing and day-to-day tech tasks
Into the MoTaverse is a podcast by Ministry of Testing, hosted by Rosie Sherry, exploring the people, insights, and systems shaping quality in modern software teams.
