Seeking Feedback on My Advanced Project “SpecPulse”

[ersinkoc]

I was inspired by this project and launched a much more advanced one called SpecPulse the GitHub repo shares the same name. I’d love to get your feedback: am I heading in the right direction? Your insights would mean a lot. Also, I believe I might be able to contribute significantly to this original project as well.

[ImBIOS]

1. I recommend you put a link, so we can save an average of 18 seconds to open Google and search manually
2. If someone creates a public YouTube video of using this, it would be more helpful to save other (busy) people time
3. Your project direction is good; the only lacking is mass testing by the community. From now on, I suggest focusing on how to get as many people to try and give you feedback on the actual project creation, maintenance, and development.

I've tested spec-kit personally to build 2 projects, and I saved and archived a bunch of feedback in my brain. It's still lacking in several critical parts. I'm testing it using some cheap models. I don't know what happened if I test using only gpt-5-codex-high, gpt-5-high, or Opus 4.1.

I believe we need to artificially create a lab-grown human brain, then we can focus on LLM after we try to train that lab-grown human brain to think, if we really want to mimic humans. That's just a belief, open to being challenged with a better belief.