Man where in the LLM training data did the “production ready “ come from? That whole list screams AI. Humans want social proof, not a list.
“describe what exists, not how to draw it”
really? please, someone tell graphviz about this, they should use it! /s
If it doesn't provide controllable layout then I might just as well generate Mermaid or DOT, which are both produced perfectly fine by LLMs.
Does address a real use-case - might be great as a library or a lightweight alternative to Mermaid.
As SaaS it's a very hard sell.
but why not just use https://github.com/jsongraph?
When doing this, do you give it a codebase and ask for a diagram, or is it more of an iterative back-and-forth where you describe the system to it.
Claude was able to determine the project, analyze the code, and infer the infrastructure. It worked through the projects as it went drawing a higher level diagram first. Then I iteratively asked for more detail around the sections I thought needed more detail.
This was just an experiment that I spent a few hours on. Months later I misplaced the mermaid markdown file so I asked Claude to reproduce the mermaid markdown from the png image so I could have it add more detail around another section that looked incorrect because it encapsulated the database and worker for one of the systems in a single box. The detail made the diagram in that area much more clear.
To me it is very good at writing/reading/reverse engineering mermaid diagrams.
Prompts like, "Please write a comprehensive report on _____ to work with _____. Include a holistic report on architecture and meaning and purpose of all involved systems. Describe the why and how of the changes in depth and include a full glossary of terms and systems. Write as a new .md in docs when you are sure there are no major gaps in your understanding. Include a report on the plan to _____."
Will be the rough shape you want to get it to dig through all the relevant code and make relevant architectural diagrams. Guide more or less towards specifics as appropriate. This has worked well since Opus 4.5.