Running Claude Code on Opus is powerful. But most turns in a session are mechanical, reading files, writing boilerplate, simple edits. You're paying top-tier rates for work that Haiku could handle in its sleep. Anthropic just released the Advisor Tool: A server-side pattern where a cheap executor model consults Opus mid-generation for strategic guidance. But it's API-only. Can't use it inside Claude Code. So I rebuilt the pattern inside Claude Code using what already exists. The setup: Opus stays in the orchestrator seat. It plans. It makes architecture decisions. It reviews output. It never touches files directly unless the task genuinely needs Opus-level judgment. Everything else gets dispatched to cheaper models: - Agent({ model: "haiku" }) — Claude subagents with full file access for simple edits - Agent({ model: "sonnet" }) — for multi-file changes that need moderate reasoning - A CLI tool (ask.py) that routes to Gemini, Kimi, MiniMax, or local Gemma via Ollama — for code generation, research, video analysis, anything where you just need text back The routing logic: Does the task need file access? → Agent tool (haiku/sonnet) Is the code complex? → sonnet or gemini Is the code simple? → haiku, kimi, or minimax (cheaper) Need video analysis? → gemini --video (native, no frame extraction) Need parallel research? → kimi-swarm (spawns up to 100 sub-agents) Want zero API cost? → gemma running locally via Ollama When a subagent hits something it can't handle, it reports back NEEDS_GUIDANCE. Opus thinks it through and re-dispatches with better context. That's the advisor pattern, strategic guidance exactly when needed, cheap execution everywhere else. Cost impact: A 10-task session where 8 tasks go to Haiku and 2 to Sonnet — your Opus tokens are only spent on planning and review. Maybe 20% of the total token volume. The other 80% runs at Haiku rates. With local models mixed in, it drops further. Open-sourced the whole thing:

I'm curious what skills you guys are installing at the user level versus the project level. I put UI UX Pro Max at the project level initially, but I found that it's applicable to most of my projects, so I just added it at the user level. Now I'm in "discovery mode" for skills and wondering what gems people have come across, especially ones that earned a place at your user level skills. Thanks in advance for sharing!

Alright, I need to run something by you all. Jake and I have been talking about building something bigger. Not just more content. Something structured. Something with real accountability. Something for anyone who wants to actually build with AI, whether you joined this community yesterday or you've been here from the start. Here's what we're considering: The Lyceum — the original Lyceum was started in ancient Greece and known as the first school of Aristotle. Ours is a 12-week program with live instruction from Jake and other AI instructors from Eduba. Small cohorts. Real projects. You'd be building something from week one, not just watching tutorials. We're thinking three different tracks: - Technical — for developers, engineers, people building tools and systems - Business — for ops people, managers, founders, consultants who need to direct AI work without necessarily writing code - Creator — for content creators, marketers, educators, solo operators building their own production systems Same core methodology across all three. Different emphasis based on what you're actually trying to do. And here's where it gets interesting. We're thinking about making it a competition. A grand champion who gets a $100K build from Eduba. First, second, third place for each cohort. Demo day at the end where people present what they built. We'd also be issuing Eduba's first-ever certification. Something backed by the same methodology we've used to train Fortune 500 teams. This is still in the planning phase. We haven't finalized everything yet. But before we lock it in, I want to hear from you. Does this sound like something you'd actually want? Drop a comment. Tell me what excites you. Tell me what concerns you. Ask questions. If there's something you'd want to see included, let me know. More details coming soon! Eduba Case Studies: https://services.eduba.io/#cases

If you’re deep in Level 3 “Building Your Stack” or just tired of wrestling with half-baked AI memory hacks, I’ve got something you’re going to love. **Meet OB1 — Open Brain** . It’s not another rabbit hole. It’s the infrastructure layer for your thinking: One database + one AI gateway + one chat channel — and any AI (Claude, ChatGPT, Cursor, Claude Code, whatever drops next month) plugs straight in. No middleware. No SaaS chains. No Zapier spaghetti. Why this matters right now (the pain we all feel) Right now “managed memories” are a vendor-lock-in trap: - Claude has its own memory. - ChatGPT has its own memory. - Cursor has its own. - Every new model that ships next quarter will have its own. Your life story, your household knowledge, your job-hunt pipeline, your taste preferences — they’re all trapped inside proprietary silos. Want to switch providers? Start from zero. Want to let two AIs talk about the same context? Build brittle integrations and pay for every hop. Want semantic recall across tools? Good luck.That’s not hacking — that’s renting brain space from landlords who can change the locks whenever they feel like it. OB1 flips the script You get one single source of truth: a Supabase Postgres table (thoughts) with built-in vector search. Every AI talks to the same persistent memory of you through an open Model Communication Protocol (MCP) gateway. - Capture thoughts via Slack/Discord/email → instantly embedded and indexed. - Any AI reads/writes with row-level security so your household data stays private. - No more copy-paste hell. No more “which model remembers what?” It’s literally the open brain layer that the big vendors refuse to give you — because once you own your memory, you own your stack. Built for us The repo is designed for exactly the kind of people in this skool: - 45-minute AI-assisted setup (Cursor or Claude Code can literally build it for you from the README). - Extensions folder with ready-to-extend modules (household KB, meal planner, CRM, job-hunt pipeline). - Primitives, schemas, recipes, and dashboards so you can fork, remix, and ship your own agent memory layers. - Full TypeScript + Python + Supabase stack — perfect for the “custom UIs + infrastructure” crowd.

you know how we're manually updating our Claude folders every session? there's a free open source tool called OpenViking that does that automatically it organises our memories, resources and skills into a file system that updates itself as we work. no more re-explaining our project every new chat — the longer we use it, the smarter Claude gets about our stuff how to connect it to Claude: we don't need VS Code for this. just run pip install openviking in the terminal, set up our Claude API key in a config file, and launch it with openviking-server. from there it reads all our files and feeds Claude the right context automatically — only pulling in what's actually relevant to what we're working on at that moment. it basically replaces the manual folder system we're building here how it helps our projects: 🧠 Claude remembers our last session without us touching anything 📁 our files stay updated automatically as we work 🔍 cleaner outputs because it only loads relevant context — not everything at once where the bag comes in 💸: once we understand this, we can sell it as a service. businesses are struggling to get AI working consistently and we'd know exactly how to fix it - charge to set up a custom memory system for clients - package it into a done-for-you AI automation service - build something useful with it and sell the template - consult — most people don't even know this exists everything we're learning in this course is literally the foundation of how this works — we're learning the right way first and then eventually we can use this tool to help us automate tasks better and even auto organize as you keep adding files and working. hope this helps our advanced users came across and thought I should share. → openviking.ai → github.com/volcengine/OpenViking