Fasyn: FastCGI Library

Hello, developers.

In 2023, I had created an experimental FastCGI implementation called Fasyn to integrate a payment server.

At the time, I read the FastCGI specification myself and implemented it piece by piece. Later, however, the need for the library disappeared, so I stopped development and left it untouched for quite a while.

Recently, I needed to connect a local MCP server to tunnel-client, so I resumed development of Fasyn.

The project is available here:

https://github.com/hodong-kim/fasyn

This time, the development process is completely different from before.

With the recent advances in AI, I redeveloped the entire Fasyn codebase in Ada, a language that works well with AI.

GPT directly read the FastCGI specification and, based on it, implemented, compiled, and tested the library.

I did not write a single line of code during this redevelopment of Fasyn.

In the past, a person read the specification and wrote the code. This time, a person set the development direction and requirements, while GPT read the Fasyn roadmap and the FastCGI specification and carried out the actual implementation and verification.

What I mainly did was write and review the roadmap. GPT helps me with even that. It seems that all I really did was decide the direction of development.

Fasyn has now been implemented and tested to the point where it can be used as a FastCGI server library.

In fact, a Sonbal local MCP server that uses the Fasyn library is already running. I also plan to release this much-anticipated MCP server soon.

FastCGI may not be used as often as it once was, but if you need a FastCGI library for Ada, I hope you will consider Fasyn as one of the options.

Thank you.