Thank you for your recognition and feedback. I quickly updated the third version. This...

Thank you for your recognition and feedback.

I quickly updated the third version.

This version is already very stable and I am quite satisfied.

It is highly recommended that everyone upgrade to this version!

(The picture in Appendix 7 is the latest effect) We have received a lot of feedback from friends, among which codex compatibility and unstable picture output are the most common ones.

Therefore, we seriously solved the problem of "how to stably produce sample-level technical drawings".

🔥 These seven latest pictures are all generated by codex ps: Claude has recently become smarter and recommends everyone to use Codex (GPT5.4 high) to produce pictures 👏 Welcome to give it a triple link and give it a 2K star: Specific update content: ✅ Codex is perfectly compatible - no more errors, and thanks to several friends for their PR ✅ Pushed to Install in one sentence: npx skills add yizhiyanhua-ai/fireworks-tech-graph One sentence upgrade: npx skills add yizhiyanhua-ai/firew orks-tech-graph --force ✅ Upgrade 7 styles from "skin-changing" to "generation system with its own layout language".

Instead of just changing the color, the title area, container, arrow semantics, legend, and brand sense are all written into the logic.

✅ Several old problems that most affect the look and feel have been severely eliminated: - Lines passing through the frame - Label pressing - The path is too mechanical - The composition is loose - The style is like a set of templates ✅ More importantly, this time it is not a few pictures that are beautifully retouched by hand.

Instead, these capabilities are abstracted into formal capabilities, basically ensuring that no errors are reported during the svg code output process: - style_overrides - header_prefix / header_text - side_label - window_controls / meta_* - blueprint_title_block If you continue to expand the image in the future, you no longer need to rely on a bunch of hardcode to support it.

This project is finally starting to smell like a "production-grade generator".

🚀 👀 Comment on this version in one sentence: It has evolved from "capable of producing pictures" to "more like a picture production system that a real team will maintain for a long time." #AI #Agent #RAG #OpenSource #Engineering #Visualization #DeveloperTools

Visual summary

Article argument map

Generated from the post's content graph

Mermaid outline

flowchart LR
  format-article["article feature"]
  topic-agents["agents"]
  topic-ai-workbenches["ai workbenches"]
  topic-retrieval["retrieval"]
  topic-tooling["tooling"]
  capability-agent-workflow["agent workflow"]
  capability-ai-native-workbench["AI-native workbench"]
  capability-developer-tooling["developer tooling"]
  capability-technical-writing["technical writing"]
  capability-product-surface["product surface"]
  format-article -->|covers| topic-agents
  format-article -->|covers| topic-ai-workbenches
  format-article -->|covers| topic-retrieval
  format-article -->|covers| topic-tooling
  format-article -->|signals| capability-agent-workflow
  format-article -->|signals| capability-ai-native-workbench
  format-article -->|signals| capability-developer-tooling
  format-article -->|signals| capability-technical-writing

Visual structure

Essay structure map

Built from summary and key paragraph positions

Mermaid outline

flowchart LR
  thesis["Thank you for your recognition and feedback. I quickly updated the third version. This version is already v..."]
  signal["Thank you for your recognition and feedback."]
  operator["🔥 These seven latest pictures are all generated by codex ps: Claude has recently become smarter and recomm..."]
  implication["🚀 👀 Comment on this version in one sentence: It has evolved from \"capable of producing pictures\" to \"more..."]
  thesis -->|frames| signal
  signal -->|develops| operator
  operator -->|lands in| implication

Source: View the original post