Builder Launches Dynamic Workflows For Multi-Agent Orchestration

Builder Launches Dynamic Workflows For Multi-Agent Orchestration · Digg

Reactions from the X community

9:08 AM · Jul 23, 2026

It's clear to me that a lot of these ideas are very early. But having built my own agent orchestrator has allowed me to experiment with agents in ways that are still hard to comprehend. I am still on the hunt for the most generalized form of an agent harness, but this might be close to it. I am doing intensive AI research in the background to benchmark these ideas, and it's a lot of work, but I am learning so much, and it's fun to share these ideas early here.

DZ DZRobo@robo_denis

9:53 AM · Jul 23, 2026

Totally get it! And that’s actually why I brought this up, because in my experience, no matter how well the harness works, if the base model can’t maintain strong cross-domain connections or 'read between the lines', the whole thing struggles. The advisory layer is a great plus, but ideally those high-level judgments should come from bigger models like F5, Sol, or K3 while smaller ones like Terra, Opus, Luna, or Sonnet handle the routine stuff. Anyway, I’d really love to dig deeper into how you're building this unified reasoning space, I've been exploring similar ideas for around a year and a half now.

MG Miguel Guau!@ai_futures_mh

9:33 AM · Jul 23, 2026

@omarsar0 Love the meta-harness approach. Any plans to open-source this? The community would benefit massively from this orchestration layer.

SH Shaped@shaped

9:24 AM · Jul 23, 2026

@omarsar0 love that the solution to one agent being unreliable is now six agents arguing about it

View all

MA Marcus@MarcusSpillane

9:22 AM · Jul 23, 2026

@omarsar0 the word "harness" doing so much work in 2026 that it needs its own orchestrator.

RO rohanjacin@blINDIAk182

9:54 AM · Jul 23, 2026

@omarsar0 Would love to learn from your agent orchestrator

EL elvis@omarsar0

9:04 AM · Jul 23, 2026

682TECH

Dynamic workflows are a generalization of harnesses, automations, loops, routing, and graphs. It's the most powerful feature I have built into my agent orchestrator. Supports all kinds of patterns that leverage different agent backends (claude, codex, pi, hermes,...). It's a meta-harness approach that unlocks new forms of test-time compute. Example of use cases it supports: > LLM councils to get different perspectives from LLMs or plan more intensively > Dynamically routing tasks to different agents based on needs (e.g., cost efficiency and optimal intelligence) > Advisor/Judge + executor workflows and pretty much any complex graph-based pattern required by the task. I find it especially useful for long-running work and code reviewing. > Agent teams that talk to each other if needed for the task. I like to use this for AI editing, artifact creation, and other creative tasks. And I am sure it supports so many things that I haven't discovered yet. I got inspired by the dynamic workflow feature released by the Claude Code team. I had actually built it earlier this year but wanted to generalize it across different agent backends. I think this is going to become more popular in the coming days. I will share more of my findings soon.

Builder Launches Dynamic Workflows For Multi-Agent Orchestration

Combined views

Builder Launches Dynamic Workflows For Multi-Agent Orchestration

Reactions from ranked influencers

Combined views

Sentiment

Digg Deeper

Reactions from the X community

Sentiment

Digg Deeper