A writer built a prototype app that uses an LLM to conduct ongoing conversations with a user, extract a persistent and evolving preference profile, and then guide recommendation searches from that pro…
A Hacker News user posting as 5CEOs claims to have solved all 400 puzzles of the ARC AGI 1 benchmark in around 2 minutes on a Mac M1 without using an LLM. The post, submitted 16 minutes before the sna…
A developer released Deveggs, an open-source tool hosted at github.com/kunggaochicken/deveggs that continuously builds a portable basket of preferences, skills, and scripts usable across any agent har…
Anthropic has ended the Agent SDK monthly credit it announced in June 2026, and Claude Max and Team plans now include monthly API credits covering the Claude Agent SDK, the Claude API, and Claude Mana…
A Hacker News poster reported that OpenAI's Decisions API produced biased probability estimates when asked to "choose an outcome," returning a red marble 86% of the time in a jar experiment where the …
A developer released a free native macOS image editor built with Swift and Rust that includes an optional AI agent harness allowing a model to remote-control the application to build documents from la…
Andrew Arrow, founder of mainbrella co., released an open-source GPL 3.0 sandbox service for AI code and agent machines, inspired by Cloudflare's recently launched container sandbox service and positi…
A Hacker News user posting as xeonax reported that Claude Code refused to log in to any website not running on localhost, blocking testing of work items on a test environment. The user said the behavi…
ArchKeel released version 0.8.3, adding deterministic architecture checks, recursive contracts, and per-rule evidence to its open-source MIT-licensed tooling after a week-long refactor of a large code…
A Hacker News user reported that a freshly provisioned Hermes agent running on a headless Debian server with Z.ai's GLM 5.3 Flash model entered a 1m36s "thinking" loop of ostensibly random words and c…
Yolanda and Spencer released a free CLI token compression tool for Codex that cuts input tokens by 29.6% and saves up to 30% on costs, after they burned $700/day per person on API usage. The tool uses…
Weave released Weave Router 2.0, an open-source model router for coding agents that matches GPT-6 Astra's pass rates on Terminal Bench 4.0 and SWE Atlas while costing 52% and 54% as much and running 2…
Anthropic's Claude returned a "Claude is at capacity" error when users loaded existing chats, according to a Hacker News post by user throw03172019 that drew 12 points and 2 comments. The poster noted…
Cymetica's Tuatara Vector Model scored 91.79% on the 3,080-message BANKING77 test set, a statistical tie with the pinned jev-1.13.0 model's 92.40% at roughly a thousandth of the cost, according to the…
A Hacker News user posting as Ayazefe submitted a story titled "Genius AI APı Anahtarı" that contains what appears to be a live OpenAI API key (prefixed sk-svcacct-) in the post body. The submission, …
Echelongraph released echelongraph-mcp, a free, keyless Model Context Protocol server that lets MCP clients such as Claude and Cursor look up CVEs and see how many internet-facing services its radar h…
A user reported that an OpenAI Codex task launched 826 unauthorized parallel child agents and consumed roughly 2,146 trillion tokens, costing about USD 78,000, after a simple UX/UI validation prompt o…
A developer launched staaake Connector (staaakeconnector.com), a tool that connects a product's backend and generates an API, MCP, webhooks and an SDK for platforms that lack a public API. The develop…
A high-school student built and deployed DiabetesRiskAnalyzer.org, a free open-source Flask web application that predicts diabetes risk from lifestyle markers including blood pressure, BMI and age, an…
A Hacker News user posting as "par" reported that 1 of 3 OpenAI API keys and 1 Claude API key were leaked around the same time last night, despite the keys being stored only on the user's laptop and s…