Building AI Applications with OGX, the Open-Source Agentic API Server

Hi all. I've been working on this since March, mostly evenings and weekends, and this is the first time I'm showing it to anyone. A bit nervous, so please be gentle. It started as my own private setup: a few prompts to plan my training. It grew into CoAIch, an open-source Claude Code plugin with 14 skills. I built most of it with Claude Code itself. It's an early preview (0.3.0, "Achilles preview") and I use it for my own training every week. A few things I learned on the way, in case they help someone building something similar: Don't let the model do the maths. My first versions let Claude work out next week's weights. They drifted: same data, different answer on a different day. Now every number comes from a small Python script with tests, and the skill files tell Claude to run the scri

i'm the maintainer of dodaios (formerly personal jarvis), an open-source ai agent app for the desktop. apache 2.0, runs on windows, macos and linux, no account and no cloud of ours between you and your models. it works with hosted models too, but this post is only about running it locally. **what it is** one app for the things ai can do on a computer: - agents that stay around: each one has its own chat, instructions, tools, memory and schedule - an agentic ide: nine coding clis side by side in real terminal panes, or one at a time as a thread - voice: your own wake phrase, live conversations, dictation into any app. speech recognition (local whisper) and voices (piper) can run fully offline - computer use, a local wiki as memory, routines, and a marketplace with 40+ plugins and mcp server

📰 The Roadmap to Mastering Agentic AI Systems in 2026 A four-level technical roadmap for building agentic AI systems, from single cognitive loops and stateful memory through multi-agent swarms and production observability. 📰 Source: KDnuggets 🔗 Link: https://www.kdnuggets.com/the-roadmap-to-mastering-agentic-ai-systems-in-2026 # AI # ArtificialIntelligence