Understand existing code
See how classes, tables, forms and extensions relate across the models you choose, so a large codebase is easier to read and explain.
Understand your X++.
Change it with context.
XppAtlas is AI-enabled code and system intelligence for Microsoft Dynamics 365 Finance & Operations, with a ready AI coding harness for your team. It helps developers, architects and AI assistants work from the real code and metadata of your project, not from guesswork.
What XppAtlas helps you do
Finance & Operations solutions combine Microsoft standard code, ISV solutions and your own extensions. XppAtlas helps your team find its way through all of it.
Find what you need
Search code and metadata, check types and methods, and follow references to the places that matter.
Investigate problems
Gather evidence with sources you can check, keep facts apart from assumptions, and see clearly what is still unknown.
Make better-informed changes
Review dependencies and extension points before you change something, so design and review decisions rest on the actual project.
An AI coding harness for D365 teams
XppAtlas comes with a ready project harness that gives AI coding assistants the rules, skills and workflow of a Finance & Operations team.
- Claude Code
- Codex
- Antigravity
One project brain for every assistant
Shared project rules and context, with ready setups for Claude Code and Codex. Antigravity can connect to the same knowledge over MCP.
Skills for everyday D365 work
Ready commands for impact analysis, extension design, Chain of Command and batch scaffolds, code review, performance fixes, tests and FDDs.
Specialist roles
Developer, architect and functional-consultant agents, plus optional paired architect and developer sessions with agreed acceptance criteria and a review loop.
Structured work, you stay in control
Each work item gets its own task folder and working notes. Changes reach your source, commits or pushes only with your explicit approval.
Use cases
A few everyday situations where project context makes the difference.
Getting to know a solution
A developer new to a project explores how a customized process is built: the classes, tables, extensions and event handlers involved, each traced back to its source.
Extending standard code
Before extending a standard method, the team looks at what already extends it and what depends on it, then chooses the extension approach with that in view.
Writing X++ with an AI assistant
The assistant looks up real tables, fields and methods in your models before it suggests code, which helps its suggestions fit your solution rather than generic examples.
Looking into an issue
Starting from the object or behaviour involved, the team follows the related code paths, collects evidence with sources and keeps open questions visible instead of guessing.
How it works
General-purpose AI assistants know X++ as a language. XppAtlas gives them, and your team, the specifics of your own solution.
Select your sources
Choose the Microsoft standard, ISV and custom models that matter to your project, from code and metadata you already have.
Build connected knowledge
XppAtlas links code, metadata and their relationships into one searchable knowledge base.
Work from real context
Developers, architects and MCP-compatible AI assistants draw on that knowledge, with references back to the source.
Your knowledge base stays under your control. What reaches an AI provider, through a connected assistant or an optional cloud service, depends on the configuration you choose.