Multi-Agent-Coder Is #12 on Stanford's TBench. Beats Claude Code
This weekend I built a multi-agent coding system which, quite unexpectedly, beat Claude Code on Stanford's Terminal Bench! The architecture is straightforward, consisting of an orchestrator agent that deploys explorer & coder subagents to complete complex terminal based tasks, utilising an intelligent context sharing mechanism along the way which makes it all work. The repo has a lot of technical details, and all the code and prompts for you to play around with if you'd like! I had a lot of fun making this, I hope you have fun reading the README, using it yourself, or even extending it! As…
Multi-Agent-Coder Is #12 on Stanford's TBench. Beats Claude Code launched on September 3, 2025 with 5 votes and more than 75% of that year's launches. The site was still online in September 2026, 1 year on.
What it does
Reached #13 on Stanford's Terminal Bench leaderboard. Orchestrator, explorer & coder agents working together with intelligent context sharing. - Danau5tin/multi-agent-coding-system
The orchestrator acts as the brain of the operation - it receives the user's task but never touches code directly. Instead, it: The orchestrator's lack of direct code access forces proper delegation and verification patterns, leading to more strategic solutions. The model uses xml tags to define an action, and YAML to define the parameters of an action if they exist. Below shows the model thinking and then dispatching a coder agent. < think > Okay, now with task_003 identifying a backup, the next step would be to attempt recovery using this identified file. </ think > < task_create > agent_type: 'coder' title: 'Attempt recovery using the identified backup file' description: | Execute the…from github.com
In the maker’s words, at launch
This weekend I built a multi-agent coding system which, quite unexpectedly, beat Claude Code on Stanford's Terminal Bench! The architecture is straightforward, consisting of an orchestrator agent that deploys explorer & coder subagents to complete complex terminal based tasks, utilising an intelligent context sharing…
Does a similar job
all alternatives →
Spotlight by BackplanesJun 2026 · backplanes.com · ▲394Hola HN! Long time lurker, sometimes commentor, first time poster here. I’ve been working alongside my two co-founders and a few colleagues on a project I’m excited (and a little nervous) to share with you all! Like many of us, I’ve lived a tortured existence with AI coding (is it vibes?) over the past few months - I think November was a big moment with this. But, one thing I’ve noticed after building orchestration layer after orchestration layer is that the thing I always came back to was “what the hell is Claude Code actually doing?” Perhaps it’s because of the time Claude Code got…
Munder DifflinAug 2026 · munderdiffl.in · ▲200Make clones with Claude Code and Codex to do your work
ReplicasJun 2026 · replicas.dev · ▲239Hey HN! I'm Connor, creator of Replicas. I'm a heavy Codex/Claude Code user, and I wanted to hand off tasks to agents like I would to an engineer - kick something off before work, let it scaffold while I'm at lunch, even come back to a ready PR. The problem: existing alternatives use workspace VMs that are black boxes. I have no access to any of them - they can't run tests, read logs, or use tools like the Playwright MCP (which I use to verify agents actually fixed what they claim to fix). Replicas gives agents real cloud workspaces where they can: - Run your test suite and read the…
20+ Claude Code agents coordinating on real work (open source)Feb 2026 · github.com · ▲53Single-agent LLMs suck at long-running complex tasks. We’ve open-sourced a multi-agent orchestrator that we’ve been using to handle long-running LLM tasks. We found that single LLM agents tend to stall, loop, or generate non-compiling code, so we built a harness for agents to coordinate over shared context while work is in progress. How it works: 1. Orchestrator agent that manages task decomposition 2. Sub-agents for parallel work 3. Subscriptions to task state and progress 4. Real-time sharing of intermediate discoveries between agents We tested this on a Putnam-level math problem, but the…
Terminal-Bench-RL: Training long-horizon terminal agents with RL2025 · github.com · ▲125After training calculator agent via RL, I really wanted to go bigger! So I built RL infrastructure for training long-horizon terminal/coding agents that scales from 2x A100s to 32x H100s (~$1M worth of compute!) Without any training, my 32B agent hit #19 on Terminal-Bench leaderboard, beating Stanford's Terminus-Qwen3-235B-A22! With training... well, too expensive, but I bet the results would be good! *What I did*: - Created a Claude Code-inspired agent (system msg + tools) - Built Docker-isolated GRPO training where each rollout gets its own container - Developed a multi-agent…
ViberiaMay 2026 · getviberia.com · ▲146Hey HN, This is my take on the agent harness. Everything on an isometric map. Agents are grouped into "buildings" that run in a sequence or a loop; e.g., the CodeForge has an agent that writes a PRD, another one that implements, and a third that reviews. Everything is customizable, you build your own buildings/teams however you want. It's a Tauri app, really light (about 8x less energy than the closest competitor I benchmarked, so it actually runs from a coffee shop on battery). It's macOS only for now, but ping me if you are willing to test the Windows or Linux version. I've been…
More ai this month
the category →
Paintings by AI models: programs run through a simulation of oil paint.
AI · 21d ago · stillwet.art
PixVerse R2▲381A real-time world model you can explore and change
AI · 30d ago · world.pixverse.video
Source: Show HN post ↗ · launch text quoted in part
Launched alongside, September 2025
the whole month →
The site hit #1 on HN about a year ago. Got enough traffic that I turned it into a proper API. What's there: - REST API — 35k products, manually filterable by category, free key, 60 req/min - CLI — npx anycrap random -c food - faker.js plugin — bundled, no API call, works offline - HuggingFace dataset — full 35k rows for ML/training https://anycrap.shop/developers
AI · 2025 · anycrap.shop
I made a built-from scratch Wayland Compositor to display any GUI app* in the terminal! I think there is a lot of unexplored potential in custom Wayland compositors, a lot of really cool things you can embed existing applications into! So, I started with embedding apps into the terminal because that is the easiest input/output (output is just utf-8 and I use the great `chafa` library for that, and I just read from stdin for the input). If you have any other ideas for cool Wayland compositors, let me know. I purposedly wrote 80% the app in Typescript to appeal to the most developers and…
Dev tools · 2025 · github.com
- IR
Years ago I stumbled across a basic version of this concept and it stuck with me. I knew if I was ever going to take on such a project, it would need to be flawless, but without coding experience it was just another idea that would never happen. By the end of 2024, as AI coding tools exploded everywhere, I finally had a way to make it real. I started from zero knowledge and spent months collaborating with AI agents as a learning experience. Every pixel and every function went through me. The AI translated what I asked for into code, but every decision was human. I didn't use existing OS…
AI · 2025 · mitchivin.com
