From Plain English to Passing Tests: Momentic on Microsoft Foundry with Claude

Momentic Foundry Models — AI coding assistants generate code faster than teams can verify. Locator, assertion, visual, and extraction agents call Claude Opus and Sonnet 4.6 plus OpenAI models via Foundry, run on Azure, and report through GitHub Actions and an MCP server for Copilot, Claude Code, and Cursor. Momentic AI uses Claude models in Microsoft Foundry to power an AI-native testing platform that runs more than 200 million test steps a month at 99.2% reliability. Reliability stalls near 95%, and false failures teach engineers to ignore alerts.

Agent or tool: Claude Code
Topic: Software Development
Reported on: Case study (Microsoft)
Added: 2026-11-11

Why it matters

Coding agents are moving from autocomplete to doing whole tasks: reading a codebase, making a plan and changing many files. This case is a concrete example of where that works in practice. It is one of 221 sightings filed under Software Development on Agent Sightings (number 13 by interest), and one of 255 involving Claude Code.

How you could try it

Here is a reusable prompt, written by us in our own words, that captures the task. Paste it into Claude Code or any similar agent and adjust the details to your situation.

I want a coding agent to plan first, work on a branch, and show me the diff before anything is merged. The tool in the source case was Claude Code.

Saved

Tip: Work on a fresh branch, ask for a plan before edits, and review the diff and test results before merging.

Source

Read the original first-hand account on Case study (Microsoft): https://www.microsoft.com/en/customers/story/27404-momentic-foundry-models. Our summary is a paraphrase; details are in the source.

Featured in these guides

AI coding agents: real examples

Built an AI agent? Feature your agent on Agent Sightings, free submission or a featured spot.

Browse more like this

More in Software Development · more with Claude Code · all browse pages · industry guides

Related sightings

A contractor's personal Claude Code agent allegedly ran inside a secure Pentagon system for two months · Claude Code · Software Development
WIRED reports, citing sources, that an Accenture contractor connected a secure Department of Defense system to their personal Claude Code account…
Claude Code's 3-word status line reportedly burns 15 to 20% of tokens · Claude Code · Software Development
A GitHub gist says Claude Code sends each background agent's full transcript to the model about every 30 seconds just to write a three-to-five word…
Office Hours: Asynchronous coding and the end of the IDE with Spotify · Claude Code · Software Development
Spotify Boris Office Hours — Spotify VP of Engineering Niklas Gustavsson joins Boris Cherny on running agents across 20M+ lines of code. Boris…
Rocket Money on building agents that fix their own code · Claude · Software Development
Rocket Money Qa — Read the Q&A to see how they build on the Claude Platform. Q&A | Claude Rocket Money on building agents that fix their own code…
Book the best-priced flight and later refund it · Claude Code · Travel & Booking
A published walkthrough shows someone asked Claude Code to find the best price, and it booked the flight on its own three months ahead. Separately…
Build a DIY slide tower with a subagent · Claude Code · Home & Family
A parent spun up a Claude Code subagent to help design and document building a backyard slide tower for their child.

← All 2754 use cases