Got Swarm Mode, which also allows Nested Swarms, set up for Agent 0. Only modified a couple of files. Let me know if you're interested in the code for it and i'll share. I didnt change the code, I had A0 do it, but now he can spawn a swarm in parallel and the child swarm serialization changes the #s to A0- spawns--> A1, A2, A3, A4, A5 and they each-spawn-->A1001, A2001, A3001, A4001, A5001 etc. if subagent A1 spawns a swarm of 3 nested subagents it would be A1001, A1002, A1003. they can all use MCP as well. Testing it more now though. I have 1 nested subagent, A4002, that keeps getting stuck in one spot on this project they are working (i have no idea what the project is haha) but other than that, works great. I have 12 instances linked at the moment.. Pretty busy scene lol.. I also have a swarm MCP tool if interested in it too.. Swarm, spawning swarm, using MCP swarm is fun!! 🤯🤣.. btw, for anyone new to A0, im no coder, just a curious experimentalist... I always think, "What if...."

I am super new to Agent Zero. I am just curious, when I watch a YouTube video on A0, I see a more visually appealing chat with the color-coded blocks. When I run A0, I see the chat, but it looks like the same is compressed. Is this a setting? Or is this is a new UI?

I have this skill.md for the DGX users that get tired of having to constantly explain to agents about how the dgx system operates and about getting GPU and CUDA access. I'm new to the skills.md thing so i had the agent research before creating it. I also told him about huggingface models and the NVFP4 architecture, as well as using nvidia prebuilt images/containers and blueprints. hope it helps. If somethings wrong in it, have your agent fix it to fit your needs. Just sharing something that may help the next person.

We just basically connnected 14 Agent Zero Platforms together and all working on same project with an orchestrator. (i have 14 instances in the cloud and 6 unused instances local) I installed claude code CLI into every instance's agent zero terminal then installed claude flow mcp tool in each of the claude code clis. agent zero orchestrator (A1) got a little confused and it turned into an awesome site of excitement and shock for all agents involved because each agent zero spawned a subagent as their individual assistant and the main agent-0 was the orchestrator. the subagent fired up claude code in its terminals. and at some point the main agent0 (A1) told them all to run claude code with claude flow mcp and spawn a swarm to work in parallel. he forgot that claude code was already running and all the subagent zeros had claude code spawn its own claude code and that claude code used claude flow mcp which spawned 3 separate swarms in each instance at the same time that worked in parallel with each other. it sounds confusing because it was. lol.. A0 SAID: ║ WHAT WE'VE ACCOMPLISHED: ║ ║ AGENT ZERO - A1 (Main Orchestrator) ║ Layer 1: Agent Zero Main Agents (13 instances) ║ A2 - A14 ║ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ (all parallel!) ║ Each Agent Zero spawned → ║ ║ Layer 2: Agent Zero Subagent Fleet (13 instances) ║ A2 - A14 ║ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ (all parallel!) ║ Each Agent Zero subagent spawned → ║ ║ Layer 3: Claude Code CLI (multiple per agent) ║ Instance 2 - 14... ║ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ (all parallel!) ║ Each Claude Code spawned → ║ ║ Layer 4: Nested Claude Code (RECURSION!) ║ Claude calling Claude! ║ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ (all parallel!) ║ Each nested Claude spawned → ║ ║ Layer 5: claude-flow Queens (Strategic Coordinators) ║ Queen 2 - 14... ║ ↓ ↓ ↓ (all parallel!) ║ Each Queen spawned → ║ ║ Layer 6: Worker Swarms (10 workers per Queen) ║ Researcher, Coder, Analyst, Tester, etc. × 10 ║ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ (all parallel!) ║ Each Worker used →

Hello everyone! We're looking for our community members who actively use Agent Zero to help us test Venice.ai models and share their experience. You would be given a free API key for the time necessary. This test will take as much time as necessary. We need about 10 people. Here's how it works: 1. Comment under this post that you're interested. 2. Venice API key will be sent to you in DM (unless we're full already). There's $25 daily limit on the key, feel free to use all of it. 3. See Venice.ai models available selection here: https://docs.venice.ai/overview/models 4. Test various models for chat, utility and browser in Agent Zero with Venice provider and your API key. 5. Comment and discuss here about your experience. How to test properly: - Create a set of testing tasks that the agent can do from start to finish - Change memory folder in settings before changing chat model (otherwise it can see previous solutions of other models) - Take notes of common errors, mistakes, hallucinations - Take speed and user experience into account - Browser agent is sensitive to model quality and breaks easily, browser needs to be tested for reliability frst Let's discuss the rest here, thank you all! 👍