I don't want to toot my own horn, but GPTScript was really just too early. So many of the patterns I was doing there have been rediscovered in Claude code, openclaw, etc. Not saying I can take credit for anything, just that my idea clearly weren't dumb, just too early.

gptscript was here too soon as well 😁 I hope this takes off ! Wonder if the issue is not the reach of your posts... Readme still states only claude support, waiting for opencode to test it.

One of these days I might go back and show people how GPTScript is actually a generic form of Claude and coding agents. Basically you can very easily build a coding agent with it. I'm not promoting that project anymore, but I feel validated in that it's ideas I guess we're good, just wrong time and wrong positioning. But if you want to look at this file you might get a bit of the idea. github.com/gptscript-ai/clio…

I realized the issue. Most devs were introduced to MCP through coding agents (cursor, VSCode) and most devs struggle to get value out of MCP in this use case. There's tons of reasons why, so they are rejecting MCP because they have a CLI and scripts available to them which are way better for them. Makes sense. Prior to MCP and coding agents I developed a thing called GPTScript that almost 100% overlaps with MCP (you won't get it if you look at it, but it does if you know the guts). I also developed a thing called Clio which was the same pattern as coding agents but before coding agents was a thing. I used to use it for coding but never realized other developers would like the pattern. And another thing called otto8 which was like a hosted version of Clio. These were all experiments and I had no idea what I was doing but in retrospect they were good ideas just too raw and early. I say all of this because the target audience for these technologies I was building was really knowledge workers. I didn't care, and still don't care about developers as end users. There are so many difficult challenges around building tools for AI for non-technical users. And I spent over a year building all these adhoc solutions for it in GPTScript. Authentication, OAuth, packaging, distribution, discovery, etc. Then MCP came around and addressed all these concerns for me. And better yet an ecosystem immediately formed with people excited to fix the issues in standard ways and I didn't have to do it all. This is great, because this type of plumbing does not differentiate a product but takes tons of time. Now I don't have to do it and can just focus on building a product. So MCP solves a lot of hard things, I know because I tried to do it myself. I totally appreciate that developers don't get it, or need it. Please reject MCP and keep working on new patterns. It helps me and the world. Because I'll just pick up those patterns and use them with MCP that solves a lot of problems you don't care about and none of the new patterns conflict with or negate MCP.

hmm, I started building something and realized that it was basically gptscript. I guess I'm building gptscript 2.0 🤷‍♂️

O3心目中的Agent平台 第一梯队：n8n、Dify、AutoGen、Flowise、CrewAI、VoiceFlow 第二梯队：LangGraph、Coze、AgentOps、GPTScript

Typescript more like Gptscript

This is gptscript from almost a year ago. Then I look at Claude Code and go, "oh, maybe I wasn't so wrong." Anyhow, working on much bigger and cooler things now.

Recent GPT domain sales: • SaudiGPT .com - $78,000 • GPTStudio .com - $6,000 • GPTScript .com - $6,000 • SearchGPT .co.uk - $6,000 • VoiceGPT .com - $5,000 Do you have any GPT domains for sale?

In this article, you will learn how to build and deploy a GPTScript agent to manage and troubleshoot a Kubernetes cluster, using natural language inputs to identify and resolve issues ➤ itnext.io/gptscript-build-yo…

@acornlabs announced Otto8 last night, & I'll be exploring that. But I've been using GPTScript to build AI Agents to perform my #DevOps tasks daily Here's a blog post about building a GCP Agent acorn.io/resources/tutorials… #DevOps #AI #LLM #OpenSource

It takes two seconds to install and use it. I'm only pitching this because it's a very eye opening experience for people to see what AI can. brew install gptscript-ai/tap/clio

In this talk, explore tools like k8sGPT or GPTScript to simplify Kyverno policy management by scanning a Kubernetes cluster and suggesting optimal policies based on best practices. hubs.ly/Q02__QnK0 #devops #security #kyverno

In this article, you will learn how to build and deploy a GPTScript agent to manage and troubleshoot a Kubernetes cluster, using natural language inputs to identify and resolve issues ➜ itnext.io/gptscript-build-yo…

GPTScript working with function calling on local Llama3.3:70B with @helixml

Finalizing the right policies to secure #K8s clusters takes a lot of manual effort🛠️ Watch @sonali_talks & @pavan_n_g at #SOSSCommunity Day to learn how to use AI tools like #clio, #k8sGPT & #GPTScript to automate policy checks to suggest optimal policies✨ @openssf

Uncover faster #AI-driven #kubernetes testing procedures with the integration of @GPTScript_ai and @Testkube_io 🤖💡 Our new tutorial shows you how to transform your testing workflow using cutting-edge #AI automation ➡️⬇️ buff.ly/3TDIrrk' #AI #LLM #DevOps #GPTScript

Started working on a new AI tutorial series. I going to learn how to become a devops :-D Not a pro of course, but I'm going to let AI teach me the basics. #clio #gptscript

Did you know our knowledge tool in @GPTScript_ai includes support for a multi-provider embedding model? You can easily manage & switch between various providers like @OpenAI, @Azure, & more using our one-time config options. Check it out here➡️⬇️ buff.ly/3WqZoWo #GPTScript #AI #LLM