Announcing Ilograph Agent Skills
Today, Ilograph is excited to release Ilograph Agent Skills, LLM agent skills for creating and maintaining Ilograph sequence diagrams of a codebase.
The diagrams generated with these skills show detailed flows of a codebase as auditable sequence diagrams:
Installation and Usage
See the linked repository above for installation and usage instructions.
Examples
 |
 |
 |
| Helm Dashboard | Vaultwarden | Zed |
What’s next
Using LLMs to generate diagrams is an exciting new frontier for Ilograph. Like all frontiers, this one has quite a bit of uncertainty and many new things to discover. How successful this will be in the long run depends on a number of assumptions about LLMs, Ilograph, and how they work together.
This is a sufficiently complex topic to warrant its own article. Next week, we’ll discuss this topic in detail in a follow-up to the now-almost-two-years-old Diagrams A.I. Can, and Cannot, Generate post.
In the meantime, if you get a chance to try out these skills on your codebase, please share your feedback on the Ilograph Discussion forum or by email at billy@ilograph.com.