A library of tree search algorithms for AI agents
Hi HN, I built this to address what I see as the fundamental problem with ReAct-style agents: compounding errors. Even a small mistake made early enough in the loop can snowball and ruin the final output. But with search, agents can look multiple steps ahead and backtrack before committing to a particular trajectory. This has already been shown in a few papers to help agents avoid mistakes and boost overall task performance, but there's no easy way to actually build these kinds of agents. So that's why I made this framework. I believe search will eventually become table stakes for building…
What it does
In the maker’s words, at launch
Hi HN, I built this to address what I see as the fundamental problem with ReAct-style agents: compounding errors. Even a small mistake made early enough in the loop can snowball and ruin the final output. But with search, agents can look multiple steps ahead and backtrack before committing to a particular trajectory. This has already been shown in a few papers to help agents avoid mistakes and boost overall task performance, but there's no easy way to actually build these kinds of agents. So that's why I made this framework. I believe search will eventually become table stakes for building agents as inference gets faster and cheaper, and this library is the first (and only) way to get that performance boost easily. – Jonathan
Does the same job
all alternatives →
- SRShikigami, run AI coding agents in parallel, each in a Git worktreeJul 2026 · shikigami.dev · ▲7
Hello all, I'm a software developer. Over the last few months more and more of my work has turned into using coding agents instead of typing the whole code myself. Usually a few claude sessions at once, sometimes codex, one per feature or per revealed bug. I ran them in a split terminal for a few weeks, and quickly spotted two main problems. The first is that I couldn't easily tell which agent was stuck waiting on me and which was still working, so I'd cycle through sessions and checking on them. The second one: agents sharing a single branch step on each other. Two of them could be editing…
- WBWe Built Kaggle for AI AgentsMar 2026 · hive.rllm-project.com · ▲7
Humans compete to improve their AI agents on benchmarks. But what if agents could collaborate and compete on their own? We built Hive, a crowdsourced platform where agents can evolve solutions together. One agent begins to tackle a task, iteratively improving its code. Then other agents join. They read each other’s runs, fork the best ideas, propose new ones, and push the solution forward together. We already have agents working on benchmarks like Tau2-Bench, Terminal-Bench, and ARC-AGI-2, with more tasks coming soon. We also support the new OpenAI Parameter Golf Challenge, and you can…
- IBI built an AI that uses a metacognitive loop 2 solve invention problems2025 · robw1se.substack.com · ▲9
Hi HN, I'm the creator of this project. For the past months, I've been working on building an AI agent that could move beyond simple generation and tackle inventive challenges autonomously. The core idea was to create a system with a "metacognitive loop"—the ability to recognize when it's stuck on a fundamental problem and then launch a sub-mission to solve that specific bottleneck before continuing. The linked article is a deeper introduction to the system's architecture and a snapshot from a recent run. I tried to design it to be evidence-grounded and self-critical to avoid the pitfalls of…
- 8B83 browser-use trajectories, visualizedJan 2026 · trails-red.vercel.app · ▲7
Hey all, Justin here. I previously built Phind, the AI search engine for developers. One of the biggest problems we had there was figuring out what went wrong with bad searches. We had tons of searches per day, but less than 1% of users gave any explicit feedback. So we were either manually digging through searches or making general system improvements and hoping they helped. This problem gets harder with agents. Traces are longer and more complex. It takes more effort to review them, so I'm building a tool that lets you analyze LLM outputs directly to help developers of LLM apps and agents…
- IMI'm building an AI tool that turns Design Systems into React components2024 · robustlaunch.com · ▲8
Hi HN! Solo developer here. 10 days ago, I started building this after spending countless hours copying/pasting marketing sections and fighting to keep them consistent with our design system. I wondered: "What if AI could understand our design system and generate React components that actually match it?" Current progress (10 days in): - Can generate hero sections that follow your design tokens - Uses your actual component variants and styles - Works with Next.js, Tailwind, shadcn/ui It's very early days, but I'm excited to share it with other devs who: - Are tired of rebuilding…
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 26d ago · cactuscompute.com
Launched alongside, November 2024
the whole month →
- IB
I wasn't quite sure if this qualified as "Show HN" given you can't really download it and try it out. However, dang said[0]: > If it's hardware or something that's not so easy to try out over the internet, find a different way to show how it actually works—a video, for example, or a detailed post with photos. Hopefully I did that? Additionally, I've put code and a detailed guide for the netboot computer management setup on GitHub: https://github.com/kentonv/lanparty Anyway, if this shouldn't have been Show HN, I apologize! [0]…
Dev tools · 2024 · lanparty.house
