AI agent for Django applications

What Merlean reads in a Django codebase

Merlean connects to your repository through GitHub or GitLab and analyses the application rather than recording a user clicking through it. In a Django project, that analysis draws on:

• The URL configuration across every installed app, including included namespaces.
• Views and viewsets, with the HTTP methods each one supports.
• Django REST Framework serializers, which state field names, types, required flags and read-only rules directly.
• Forms and model forms, including clean methods where cross-field rules live.
• Permission classes and the queryset filtering that scopes a user to their own data.
• Models and their relationships, for resolving references between objects.

The result is a machine-readable model of what the application can do, which fields each action needs, which order multi-step work has to run in, and who is allowed to run it. That model is what user requests are resolved against.

Authentication and permissions

Merlean authenticates as the signed-in user through your existing session or token authentication. Permission classes and object-level checks run untouched, and queryset scoping continues to limit what is visible.

Worth knowing about Django specifically

• DRF projects produce the most accurate models, because serializers make the contract explicit. Function-based views returning hand-built JSON require more inference.
• Permission logic implemented by filtering a queryset rather than by raising is understood, but it is worth confirming: an action that silently returns nothing reads differently from one that refuses.
• Signals that fire on save are noted, so side effects triggered by a write are represented in the workflow rather than appearing as unexplained changes.

Getting started

Connect the repository, run the first analysis, and review what Merlean understood. Reviewing is the step worth spending time on: the analysis is derived from your code, so where it is wrong it is usually pointing at something genuinely ambiguous in the application.

Then add the widget with a single script tag, or expose the application over the Model Context Protocol so an AI agent your users already use can reach it.

Frequently asked questions

Does Merlean need a Django app installed?

No. It reads the repository and calls your existing views over HTTP. There is nothing to add to INSTALLED_APPS.

Does it work without Django REST Framework?

Yes, though DRF projects analyse more precisely because serializers declare the contract. Plain Django views work; the model just leans more on forms.

How is the Django admin treated?

It is read as a source of understanding about models, but Merlean executes through your application's own views rather than through admin, so admin-only privileges are never used.

Related