The Ultimate MCP Client
Over the past couple weeks, l've been really immersed in learning about MCP, a new protocol for equipping any LLM with a set of tools that can run on your own machine or a remote server you control and give all kinds of superpowers to Al agents to do things like search, etc. As part of that research, l've already built one very fleshed-out and useful MCP server that l've shared here (I've added much more to it recently though!), the LLM Gateway MCP Server, which lets you use a big model to delegate to a cheaper model (and many more things in addition to that, like running automated…
What it does
In the maker’s words, at launch
Over the past couple weeks, l've been really immersed in learning about MCP, a new protocol for equipping any LLM with a set of tools that can run on your own machine or a remote server you control and give all kinds of superpowers to Al agents to do things like search, etc. As part of that research, l've already built one very fleshed-out and useful MCP server that l've shared here (I've added much more to it recently though!), the LLM Gateway MCP Server, which lets you use a big model to delegate to a cheaper model (and many more things in addition to that, like running automated multi-round LLM Tournaments, which I also posted about recently on X). To actually use these MCP servers though, you need an MCP client. Most people seem to be using the Claude Desktop app. I tried this and got it to work just fine, but it was a bit annoying to set up and there were lots of things I didn't like about it. I wanted something better. So two days ago I began work on what I call the Ultimate MCP Client. After ~24 hours of work, it's working and ready and I'm really proud of how amazingly well it turned out. This is going to be a workhorse tool for me personally. It's pure python and all in a single large .py file which can be deployed as a self-contained uv script if you want. It offers all kinds of features and very rich console output for use interactively in a terminal, along with a CLI. But it can also be used in the background. That kind of background functionality, orchestrating and coordinating several MCP servers nicely, is how I mostly intend on using it. But once I saw how nice the interactive terminal experience was, I realized that I could slap a FastAPl server on top of it and make a web GUl. Because I hate unneeded complexity so much, I made the WebGUl a single self-contained HTML file you can just open in your browser (similar to my Your-Source-to-Prompt tool), and it looks awesome using Alpine and Daisy and other nice Ul libraries, all loaded via CDN.
Does the same job
all alternatives →- MAMy AI Native Resume2025 · ai.jakegaylor.com · ▲301
I've been deeply involved in working with AI agents and large language models (LLMs) for a while now. During a recent job search, I found myself repeatedly explaining my skills and experiences to various assistants. Around the same time, I was creating content for my website to help hiring teams understand my capabilities better and make informed decisions. MCP had started to gain momentum and I saw a way to reduce my toil. So I built an MCP server that can effectively communicate my qualifications as a job candidate. This server acts as an AI-powered resume, providing an understanding of my…
MCP-Builder.ai11d ago · mcp-builder.ai · ▲174The fastest way to connect your data with your AI Tools.
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 16d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 26d ago · cactuscompute.com
Launched alongside, April 2025
the whole month →- IB
Hi everyone, I built PyXL — a hardware processor that executes a custom assembly generated from Python programs, without using a traditional interpreter or virtual machine. It compiles Python -> CPython Bytecode -> Instruction set designed for direct hardware execution. I’m sharing an early benchmark: a GPIO test where PyXL achieves a 480ns round-trip toggle — compared to 14-25 micro seconds on a MicroPython Pyboard - even though PyXL runs at a lower clock (100MHz vs. 168MHz). The design is stack-based, fully pipelined, and preserves Python's dynamic typing without static type restrictions.…
Dev tools · 2025 · runpyxl.com
- UC
Life & fun · 2025 · filiph.github.io
- IB
https://the-pocket.github.io/Tutorial-Codebase-Knowledge/
AI · 2025 · github.com
