- superdesigndev/treg
github-trending
OpenRouter for agent tools. Join community here: https://discord.gg/6mQYYfFMAn https://treg.to Treg (OpenRouter for Tools) OpenRouter, but for agent tools instead of models. Point an agent at one base URL with one token and it can do the job: 3,000+ catalogued endpoints across 60+ providers — SEO and backlinks, social and trends, people and company enrichment, ads, scraping, image and video generation — priced per call, from a cent, with no provider signup. Plus your own team's keys, skills and CLIs, callable by every teammate's agent without the credential ever leaving the server. Ask for the task, not the tool. You do not need to know which vendor sells backlink data, or to hold an account with them. Search for what you want to do, read the price, call it. Built for the Superdesign team…
- dream-num/univer
github-trending
The Office Harness for AI Agents — Spreadsheets, Docs, Slides, Canvas, Relational Tables, and PDF in one runtime. https://docs.univer.ai The Office Harness for AI Agents Spreadsheets · Documents · Presentations · Bases · Boards · PDFs (coming soon) High-performance, fully customizable Office SDK Build embeddable productivity experiences with a plugin architecture, Canvas-based rendering, a formula engine, and one Facade API that works in the browser and on Node.js. English | 简体中文 | 繁體中文 | 日本語 | 한국어 | Español 🌐 Website | 📖 Documentation | ✨ Showcase | 📘 API Reference | 📝 Blog ✨ What is Univer? Univer is an open-source SDK for creating office applications inside your own product. It gives you the building blocks for spreadsheet, document, and presentation experiences without forcing you…
- google/ax
github-trending
Google's open agentic orchestration runtime https://agentexecutor.io AX Warning We are still actively refining our core concepts, protocols, and specifications. We will likely to introduce major breaking changes prior to a stable release. Declare an agentic task with workspaces and gateway specifications. AX sandboxes it, wires up its workspace, fences its network, and helps running it at scale. AX is a high-throughput, declarative orchestrator to run billions of autonomous agent workloads in a cluster. It runs on top of Agent Substrate for sandboxed execution and is built to run billions of tasks per cluster. If you have used Kubernetes, ax will feel similar. # task.yaml apiVersion: ax.io/v1alpha1 kind: Workspace metadata: name: golang spec: git: - repo: https://github.com/golang/go.git …
- agent-substrate/substrate
github-trending
Agent Substrate: the core system Agent Substrate NOTE: This is not an officially supported Google product. This project is not eligible for the Google Open Source Software Vulnerability Rewards Program. What is Agent Substrate? Agent Substrate is a secure-by-default agent execution runtime engineered to run millions of sandboxes with 10x higher density than standard container runtimes. Purpose-built for the era of autonomous agents, Substrate delivers sub-500ms resume operations at over 500 suspend/resume activations per second with native zero-trust kernel and network isolation. It supports multiple sandbox technologies including microVMs and gVisor, enabling consistent lifecycle operations for all sandbox types. At its core, Agent Substrate maps a larger set of “actors” (applications su…
- mvt-project/mvt
github-trending
MVT (Mobile Verification Toolkit) helps with conducting forensics of mobile devices in order to find signs of a potential compromise. https://mvt.re Mobile Verification Toolkit Important We recently merged the "v3" branch. This introduced breaking changes. If you relied on mvt output in other scripts They might have broken. More details: https://github.com/mvt-project/mvt/issues/757 Mobile Verification Toolkit (MVT) is a collection of utilities to simplify and automate the process of gathering forensic traces helpful to identify a potential compromise of Android and iOS devices. It has been developed and released by the Amnesty International Security Lab in July 2021 in the context of the Pegasus Project along with a technical forensic methodology. It continues to be maintained by Amnesty…
- browser-use/video-use
github-trending
Edit videos with coding agents video-use Introducing video-use — edit videos with Claude Code. 100% open source. Drop raw footage in a folder, chat with Claude Code, get final.mp4 back. Works for any content — talking heads, montages, tutorials, travel, interviews — without presets or menus. Try video-use in Browser Use Cloud. What it does Cuts out filler words (umm, uh, false starts) and dead space between takes Auto color grades every segment (warm cinematic, neutral punch, or any custom ffmpeg chain) 30ms audio fades at every cut so you never hear a pop Burns subtitles in your style — 2-word UPPERCASE chunks by default, fully customizable Generates animation overlays via HyperFrames, Remotion, Manim, or PIL — spawned in parallel sub-agents, one per animation Self-evaluates the rendered…
- davila7/claude-code-templates
github-trending
CLI tool for configuring and monitoring Claude Code https://aitmpl.com Sponsored by Bright Data Bright Data proudly supports this project. Connect Claude Code to live web data — try their Skills and MCPs for web search, scraping, and structured data feeds. # Install the Bright Data Skills + MCP for Claude Code npx claude-code-templates@latest --skill web-data/search,web-data/scrape,web-data/data-feeds,web-data/bright-data-mcp,web-data/bright-data-best-practices,development/brightdata-local-search --mcp web-data/brightdata --yes Claude Code Templates (aitmpl.com) Ready-to-use configurations for Anthropic's Claude Code. A comprehensive collection of AI agents, custom commands, settings, hooks, external integrations (MCPs), and project templates to enhance your development workflow. Browse &…
- How to Create Your Own Personal AI Benchmark
hn-ai· 23-sept
Article URL: https://every.to/also-true-for-humans/how-to-create-your-own-personal-ai-benchmark Comments URL: https://news.ycombinator.com/item?id=49811036 Points: 2 # Comments: 0
- Created a local AI to save data
hn-ai· 23-sept
Article URL: https://github.com/just-not-google/BiNeuron Comments URL: https://news.ycombinator.com/item?id=49810817 Points: 1 # Comments: 0
- Show HN: Visually orchestrate Claude Code AI agents
hn-ai· 23-sept
Article URL: https://github.com/rondoflow/rondoflow Comments URL: https://news.ycombinator.com/item?id=49812163 Points: 3 # Comments: 0
- Show HN: use-citations – Claude Code skill for citation backed answers
hn-ai· 23-sept
This is a Claude Code skill for research questions in law, tax, medicine and similar fields, or for anyone wanting to ask knowledge-based or research questions while in a Claude Code shell really. It gets claude to save the source documents, answers from them, and builds an artifact/report where each sourced phrase opens in a slide out drawer with the relevant passage highlighted in the full source doc (quick video demo of the artifact is in the linked repo). Tools like Harvey, Legora, and OpenEvidence provide a lot more value than just this, so it isn’t a replacement. I imagine, though, that there are some people in those fields around the world who might have a Claude/GPT subscription and not one of the domain specific AI tools. I know CLI might not be natural to all in the those indust…
- Big Banks Say They're Uneasy About People Shopping via AI Agents
hn-ai· 23-sept
Article URL: https://gizmodo.com/big-banks-say-theyre-uneasy-about-people-shopping-via-ai-agents-2000815443 Comments URL: https://news.ycombinator.com/item?id=49811556 Points: 9 # Comments: 2
- A Summer of AI Optimization
hn-ai· 23-sept
Article URL: https://lemire.me/blog/2026/09/22/a-summer-of-ai-optimization/ Comments URL: https://news.ycombinator.com/item?id=49812138 Points: 2 # Comments: 0
- Mastering LLM Inference Optimization
hn-ai· 23-sept
Article URL: https://machinelearningmastery.com/the-roadmap-to-mastering-llm-inference-optimization/ Comments URL: https://news.ycombinator.com/item?id=49812210 Points: 2 # Comments: 0
- New AI technique could make minimally invasive surgeries safer and more precise
hn-ai· 23-sept
Article URL: https://news.mit.edu/2026/new-ai-technique-could-make-minimally-invasive-surgeries-safer-more-precise-0916 Comments URL: https://news.ycombinator.com/item?id=49812778 Points: 2 # Comments: 1
- Ask HN: How do you cope with AI being used to target and bomb children?
hn-ai· 23-sept
Comments URL: https://news.ycombinator.com/item?id=49812905 Points: 2 # Comments: 1
- America is in the wrong AI race with China
hn-ai· 23-sept
Article URL: https://restofworld.org/2026/america-china-ai-race-trust/ Comments URL: https://news.ycombinator.com/item?id=49811640 Points: 6 # Comments: 0
- Claude Opus 5.5 – Pelican Game
hn-ai· 23-sept
Article URL: https://claude-opus-5-5.riba2534.cn/ Comments URL: https://news.ycombinator.com/item?id=49812241 Points: 4 # Comments: 2
- Measuring behavioral signals of LLM through psychometric profiling
hn-ai· 23-sept
Article URL: https://arxiv.org/abs/2609.22934 Comments URL: https://news.ycombinator.com/item?id=49811024 Points: 1 # Comments: 0
- Specs cut defects in AI-generated code from 148 to 23 across five models
hn-ai· 23-sept
Article URL: https://arxiv.org/abs/2609.23270 Comments URL: https://news.ycombinator.com/item?id=49811333 Points: 1 # Comments: 0
- Emergent Collusion in Long-Horizon LLM Agent Interaction
hn-ai· 23-sept
Article URL: https://arxiv.org/abs/2609.24967 Comments URL: https://news.ycombinator.com/item?id=49811171 Points: 1 # Comments: 0
- Recursive self-improvement of AI research agents
hn-ai· 23-sept
Article URL: https://arxiv.org/abs/2609.26457 Comments URL: https://news.ycombinator.com/item?id=49812641 Points: 1 # Comments: 0
- Claude Code Hooks Explained: The Deterministic Layer Around Your Agent
hn-ai· 23-sept
Article URL: https://blakecrosley.com/blog/claude-code-hooks-explained Comments URL: https://news.ycombinator.com/item?id=49812522 Points: 1 # Comments: 0
- Microsoft disrupts AI-assisted platform that compromised 12,000 accounts
hn-ai· 23-sept
Article URL: https://arstechnica.com/security/2026/09/microsoft-disrupts-ai-assisted-platform-that-compromised-12000/ Comments URL: https://news.ycombinator.com/item?id=49812609 Points: 6 # Comments: 0
- Which AI model should your team use for Ansible
hn-ai· 23-sept
Article URL: https://steampunk.si/spotter/ai-leaderboard/ Comments URL: https://news.ycombinator.com/item?id=49812391 Points: 2 # Comments: 0
- Tradeoff considerations while running LLM models locally
hn-ai· 23-sept
Article URL: https://swapniltalekar.substack.com/p/should-you-be-running-llm-models Comments URL: https://news.ycombinator.com/item?id=49812956 Points: 1 # Comments: 0
- Snorkel AI triples valuation to $3.5B as demand for AI training data booms
hn-ai· 23-sept
Article URL: https://techcrunch.com/2026/09/22/snorkel-ai-triples-valuation-to-3-5b-as-demand-for-ai-training-data-booms/ Comments URL: https://news.ycombinator.com/item?id=49811361 Points: 2 # Comments: 2
- Benzi – Code Intelligence MCP/Harness beating Claude Code and CodeGraph
hn-ai· 23-sept
Article URL: https://varianttech.net Comments URL: https://news.ycombinator.com/item?id=49812151 Points: 2 # Comments: 1
- Show HN: Ceptile AI Search platform which combines web results with AI chat
hn-ai· 23-sept
I built Ceptile because I wanted a search tool that could combine the convenience of an AI answer with the ability to inspect the actual web pages behind it. A Ceptile search retrieves pages from the web, ranks the retrieved sources, and uses them to produce an answer with links back to the sources. You can open the sources and continue researching from there. I've also been building some things around the search workflow, including Notes for keeping research, translation tools, and ways to work with information found during a search. Just select a text in answer or web results, and it will show a toolbar to save it to notes, translate, etc. The project is still early. A lot of the work so far has been figuring out the search pipeline itself: retrieving useful pages, dealing with noisy re…
- EN World Is Being DDoS'd by AI Bots
hn-ai· 23-sept
Article URL: https://www.enworld.org/ Comments URL: https://news.ycombinator.com/item?id=49810493 Points: 2 # Comments: 1
- The 'AI confidence trap' many New Zealanders are caught in
hn-ai· 23-sept
Article URL: https://www.rnz.co.nz/news/business/1530184/the-ai-confidence-trap-many-new-zealanders-are-caught-in Comments URL: https://news.ycombinator.com/item?id=49810684 Points: 8 # Comments: 0
- A Machine Beat Me at Chess. I'm Still Rooting for AI
hn-ai· 23-sept
Article URL: https://www.thenextmove.org/p/a-machine-beat-me-at-chess-im-still Comments URL: https://news.ycombinator.com/item?id=49811408 Points: 1 # Comments: 0
- Naming AI agents after Seinfeld characters helps bots join the team – report
hn-ai· 23-sept
Article URL: https://www.theregister.com/ai-and-ml/2026/09/22/security-firm-finds-naming-ai-agents-after-seinfeld-characters-helps-bots-join-the-team/5298424 Comments URL: https://news.ycombinator.com/item?id=49810295 Points: 3 # Comments: 0
- United Nations: AI Agents, Misalignment and the Risk of Losing Human Control [pdf]
hn-ai· 23-sept
Article URL: https://www.un.org/independent-international-scientific-panel-ai/sites/default/files/2026-09/Thematic%20Brief_AI%20Agents%2C%20Misalignment%20and%20the%20Risk%20of%20Losing%20Human%20Control_Evidence%20from%20the%20OpenAI-Hugging%20Face%20Incident_Independent%20International%20Scientific%20Panel%20on%20AI_Advance%20Unedited%20Version%201_21%20Sept%202026.pdf Comments URL: https://news.ycombinator.com/item?id=49811485 Points: 1 # Comments: 0
- Claude Pop – I'm Upping My P(Doom) [video]
hn-ai· 23-sept
Article URL: https://www.youtube.com/watch?v=8j-hR4fJywU Comments URL: https://news.ycombinator.com/item?id=49811778 Points: 2 # Comments: 0
- [AINews] Claude Opus 5.5, the new default model for AINews — and everybody cuts prices 40-50%
latentspace· 23-sept
overshadowing more efficient GPT6 models from OpenAI
- [AINews] Xiaomi MiMo-V2.6-Pro 1T-A42B: the new top Open Weights model, trained for $3M
latentspace· 22-sept
crowning a new Chinese frontier lab
- Jev: System One models for Prod, not God — with Diogo Almeida, CEO, TypeSafe AI
latentspace· 21-sept
The definitive Jev podcast with its lead creator.
- 🔬 An Oscar, Two Asteroids, and the Algorithm in Your sklearn: John Platt on AI for Science
latentspace· 22-sept
We talked to Google’s Oscar winning “Giganerd” about automating science, solving climate change, and how future generations can contribute to science in the age of superintelligent AI
- Jev introduces a new shape of LLM - System One, aka Decision Models
simonw· 21-sept
<p>Last week <a href="https://typesafe.ai/">TypeSafe AI</a> unveiled <a href="https://typesafe.ai/blog/introducing-system-one-models-and-jev">Jev</a>, their first example of a new category of model that they are calling "System One models" (I'm with Maggie Appleton, I think "decision models" is <a href="https://twitter.com/Mappletons/status/2101560333441610133">a better name</a> for these). Jev is an interesting variant on the usual LLM format: it still accepts text inputs, but instead of text output it returns floating point numbers corresponding to categories, yes/no questions, ratings, and associated confidence scores.</p> <p>TypeSafe describe Jev like this:</p> <blockquote> <p>Think of Jev as a frontier-intelligence function call: unstructured state in, typed probabilistic decisions o…
- llm-anthropic 0.29
simonw· 22-sept
<p><strong>Release:</strong> <a href="https://github.com/simonw/llm-anthropic/releases/tag/0.29">llm-anthropic 0.29</a></p> <p>Adds support for <a href="https://www.anthropic.com/claude-opus-5-5">Claude Opus 5.5</a>:</p> <pre><code>llm -m claude-opus-5.5 "prompt goes here" </code></pre> <p>Tags: <a href="https://simonwillison.net/tags/llm">llm</a>, <a href="https://simonwillison.net/tags/anthropic">anthropic</a></p>
- llm-typesafe 0.1a0
simonw· 22-sept
<p><strong>Release:</strong> <a href="https://github.com/simonw/llm-typesafe/releases/tag/0.1a0">llm-typesafe 0.1a0</a></p> <p>I built this new plugin for <a href="https://llm.datasette.io/">LLM</a> to add support for <a href="https://simonwillison.net/2026/Sep/21/jev/">TypeSafe AI's new Jev model</a>. Install it like this:</p> <pre><code>llm install llm-typesafe </code></pre> <p>Then set an API key (<a href="https://console.typesafe.ai/">get one here</a>, the waitlist seems to move pretty fast):</p> <pre><code>llm keys set typesafe # Paste key </code></pre> <p>And now you can ask yes/no "noul" questions like this:</p> <pre><code>llm -m jev 'Please refund my last payment.' \ -s 'Does this message explicitly request a refund?' </code></pre> <p>Output:</p> <pre><code>{"type": "noul", "noul"…
- llm 0.36
simonw· 22-sept
<p><strong>Release:</strong> <a href="https://github.com/simonw/llm/releases/tag/0.36">llm 0.36</a></p> <blockquote> <ul> <li>New OpenAI models: <code>gpt-6-sol</code> for <a href="https://developers.openai.com/api/docs/models/gpt-6-sol">GPT-6 Sol</a> and <code>gpt-6-luna</code> for <a href="https://developers.openai.com/api/docs/models/gpt-6-luna">GPT-6 Luna</a>. <a href="https://github.com/simonw/llm/issues/1702">#1702</a></li> <li>Model plugins can now declare <code>supports_conversation = False</code> for models that only accept single-turn prompts. LLM raises <code>llm.ConversationNotSupported</code> when these models receive assistant or tool history, and <code>llm chat</code> rejects them before starting a session. See <a href="https://llm.datasette.io/en/stable/plugins/advanced-mo…
- Claude Opus 5.5, GPT-6 Sol, GPT-6 Luna, and a new price war
simonw· 22-sept
<p>Yesterday was <a href="https://x.ai/news/grok-4-7">Grok 4.7</a> (<a href="https://news.ycombinator.com/item?id=49788838#49790209">pelicans</a>) and <a href="https://mimo.xiaomi.com/mimo-v2-6">MiMo v2.6 Flash/Pro</a> (<a href="https://news.ycombinator.com/item?id=49792730#49793480">more pelicans</a>). Today Anthropic <a href="https://www.anthropic.com/claude-opus-5-5">released Claude Opus 5.5</a>, and around an hour later OpenAI <a href="https://openai.com/index/introducing-gpt-6-sol-and-luna/">released GPT-6 Sol and GPT-6 Luna</a>. It's going to take a while to get a good read on all of these new models, but here are my impressions so far.</p> <h4 id="gpt-6-sol-and-luna-are-half-the-price-of-their-gpt-5-6-equivalents">GPT-6 Sol and Luna are half the price of their GPT-5.6 equivalents</…
- Quoting @therealcornpop
simonw· 22-sept
<blockquote cite="https://www.tiktok.com/@therealcornpop/video/7687243097801051422"><p>Hey, you know it's like super obvious if you're using AI to write your scripts for TikTok and YouTube, right? [...] It's not just the general AI-isms of "it's not X, it's Y", or the rule of three, or the really weird broken staccato-like way of writing where you just say a lot of things with all these punctuation marks. and it sounds really deep, but it's not.</p> <p>It's the lack of <em>anything</em>. It's the lack of a definitive sort of spear of your voice. It's the fact I can tell you don't have opinions about the thing that you're talking about.</p></blockquote> <p class="cite">— <a href="https://www.tiktok.com/@therealcornpop/video/7687243097801051422">@therealcornpop</a>, on TikTok</p> <p>T…
- SF October 14th: A Birds of a Feather Session on Agentic Engineering
simonw· 23-sept
<p><strong><a href="https://luma.com/vxuiuyvg">SF October 14th: A Birds of a Feather Session on Agentic Engineering</a></strong></p> I'm hosting an evening event with Jesse Vincent in San Francisco on Wednesday 14th October for people who are building weird and interesting things with and on top of coding agents.</p> <p>Think of it as an agentic show-and-tell:</p> <blockquote> <p>Compare notes with other builders and experimenters on things you’re trying, what you're learning, and what you haven’t figured out yet. We’re especially interested in work you haven’t discussed publicly, odd experiments, or unfinished projects that don’t have an obvious market.</p> <p>Expect one flowing conversation with an informal show-and-tell. Sharing something you’re working on is encouraged but no presen…
- MCP was always a bad idea?
simonw· 20-sept
<p><a href="https://news.ycombinator.com/item?id=49779329#49779718">My comment</a> on <a href="https://news.ycombinator.com/item?id=49779329">MCP was always a bad idea?</a> — Hacker News.</p><p>This article entirely misses the value that MCP brings today.</p> <p>Sure, there's almost no reason to use MCPs if you are running a full-blown terminal agent (Claude Code, Codex, Meta Muse, OpenClaw etc) with unfettered internet access - just let it call APIs directly.</p> <p>If you want to operate something that's less YOLO than that, you'll find yourself wanting:</p> <ol> <li>Control over exactly which external services it can access</li> <li>A way to handle authentication that doesn't allow the agent to directly access API keys</li> <li>A sensible UI to allow users to connect and authenti…
- Cloudflare Python Workers are now generally available
simonw· 21-sept
<p><strong><a href="https://blog.cloudflare.com/python-workers-ga/">Cloudflare Python Workers are now generally available</a></strong></p> After a two year preview, Cloudflare's support for running Python code in their server-side Workers platform is now stable: "Python is now a first-class, fully supported language on the Cloudflare Developer Platform".</p> <p>A neat thing about this is how it works. Cloudflare are running Python compiled to WebAssembly via Pyodide in their V8-based <a href="https://github.com/cloudflare/workerd">workerd</a> runtime.</p> <p>This comes with some limitations, <a href="https://developers.cloudflare.com/workers/languages/python/stdlib/">documented here</a> - most notably both <code>multiprocessing</code> and <code>threading</code> are non-functional in the W…