A Show HN post introduces an AI acne coach named Ally that is described as helping users understand their acne and find its root cause. The post states that Ally is not a doctor and can make mistakes, and advises talking to a doctor or pharmacist for medical decisions.
Hugging Face released @huggingface/kernels, a minimal JavaScript library for loading and running optimized WebGPU kernels from the Hugging Face Hub, alongside an initial collection of 207 Apache-2.0 licensed kernels at huggingface.co/webgpu-kernels. Each kernel is published as a versioned repository containing a manifest, correctness and benchmark cases, and parameterized WGSL shader templates, and the team also launched Fleet, an in-browser GPU benchmarking suite that crowdsources correctness and performance evidence. In comparisons against ONNX Runtime WebGPU on an Apple M4 GPU across 809 matching cases, the kernels were 2.57x faster by geometric mean and 1.90x faster at the median.
Zumungo is a home automation tool accessed by pointing a browser at https://zumungo.local after installation; it advertises itself on the LAN via mDNS so the address resolves on Mac, Windows, iPhone, iPad, and Android, though older Linux may need Avahi installed. Because it uses a self-signed HTTPS certificate by default, browsers show a security warning that users dismiss via "Advanced → Proceed anyway." On first visit it offers to install a trusted certificate, or users can set up Cloudflare Tunnel for a real domain with a valid cert.
AWS has released an official emulator for running tests against a local clone of its services without provisioning a paid API key. It is written in Rust and accumulated 29,684 commits in its first three weeks, with Swaminathan Sivasubramanian describing its origins in a guest post on Werner Vogels' blog. It currently requires Docker on macOS and Windows, and the author hopes a single downloadable binary will follow.
Google launched the Gemini Decisions API, powered by the new Gemini Decision 1 model, which does not generate text but scores up to 1,000 typed questions against one shared input and returns calibrated probabilities for uses like routing, policy checks, tool selection and grading. It supports text, images, audio, video and PDFs with a 262,144-token context, bills $0.10 per 1M input tokens with zero output tokens, and is available only via the Google-hosted Gemini Developer API (v1beta). Google reports it leads 7 of 9 accuracy benchmarks in its table with a top-1 ECE of 0.0250, though it trails TypeSafe Jev 1.14 on When2Call (79.9% vs 82.0%) and costs more across the shared suite ($0.88 vs $0.68).
Scout is a Windows and Linux coding agent that pools the free tiers of five providers plus Kilo Code, OpenCode and OVH Cloud, falling back to the next model when one is rate limited, with free models limited to 10/min and 200/day without an account. It offers five permission modes, context pruning at 60k and compaction at 90k, self-maintaining memory in committable .scout/memory.md and ~/.scoutcode/memory.md files, and a Rust desktop app with a React WebView2 window shipping as an 18 MB executable. It also includes an issue tracker, remote control from a browser or phone, and an MCP bridge letting claude.ai, ChatGPT, Gemini and Grok use local file and shell tools.
EtherPK 0.22.0 adds automatic version history for every page and journal entry in a synced graph, creating versions as you write and before actions like delete, rename, merge, Replace all, or removing protection. Versions and deleted items are kept for a set period (30 days in the examples), thinned as they age, encrypted with the graph's key before upload, and restorable from a Version history tab or the Deleted documents and files section. AI agents connected to a synced graph can list and read versions via the list_versions and read_version tools but cannot restore them, which a person does in EtherPK.
HarborIQ offers a free CRM for independent insurance agencies covering clients, pipeline, renewals and follow-ups, designed to sit alongside an existing AMS, with CSV imports from systems such as Epic, EZLynx, HawkSoft, NowCerts and AMS360. Its paid Intelligence tier is $49 a month for the whole agency and answers carrier questions from the agency's own carrier guides, citing the source page, with 300 shared checks a month. The free CRM requires no card and has no per-seat fee.
A Hacker News submission notes that Gemini Decision 1 is available for free on Google AI Studio, linking to the AI Studio Decisions API documentation. The post had 6 points and no comments at the time of collection.
Zumungo is a free, locally hosted home automation platform that runs on hardware you already own, such as a Raspberry Pi or desktop PC, and supports over 50,000 devices from more than 3,800 brands. It is built around an AI assistant that can answer questions, control devices by voice or screen, and design events and panels, using either an online service (Google Gemini, Anthropic Claude or OpenAI) billed directly by that provider, or a fully local AI on a computer with an NVIDIA graphics card with 16 GB of memory or more. Setup is point-and-click: draw rooms, discover and drag devices into place, then control the home by talking to it.
Airlock is a Go CLI that checks AI models before they enter an air-gapped network and again before they run, and it works alongside Zarf as a deploy step. The project is available at airlock.sammatrix.in and on GitHub at github.com/samarthpatel24/airlock.
Abliteration offers an OpenAI-compatible API (base_url https://api.abliteration.ai/v1) for accessing models for security and safety work that mainstream AI often refuses, letting users keep their OpenAI or Anthropic client by changing the endpoint, API key, and model. It also describes a Policy Gateway that refuses requests outside organization-defined rules before they reach the model, and a Training Data service that turns a brief into labeled examples, preference pairs, and evaluation sets. The page states that prompts and outputs aren't stored by default or used to train models.
Google DeepMind unveiled AlphaProof Code, an AI system that mathematically proves software meets its security specifications, and demonstrated it on Signal's encryption protocol, completing in four weeks work that typically takes years. After experts defined formal security requirements, AI agents built on Gemini Carbon wrote over six million lines in the Lean proof language, flagging roughly one percent of the codebase for human review, and two deliberately planted bugs were caught immediately. Signal CTO Ehren Kret says the approach strengthens trust in data privacy as AI threats grow, and DeepMind plans to extend the method to power grids and medical devices through systems like CodeMender.
Featured coverage of this story: 《Introducing AlphaProof Code: a breakthrough in verifying complex software at depth》
llm 0.37.2 fixes a security issue reported by Daniel de Araujo in GHSA-v9h7-jj4g-f862 and GHSA-r639-rr7r-x72w. The extra-openai-models.yaml mechanism, now managed by the new-in-0.37 `llm openai endpoint add` command, could accidentally send your configured OpenAI API key to another provider if the key named by api_key_name had never been stored, or if a `responses: true` model was run with `-o chat_completions 1`.
Simon Willison added tag-based browsing to his collection of 244 vibe-coded tools, using tree-sitter to analyze each tool's code for JavaScript APIs, libraries, and web API hosts. The homepage now features tag clouds for browsing by each of those tag categories.
Google introduced the Gemini Decisions API, powered by the new Gemini Decision 1 model (gemini-decision-1) and available through the Gemini API in Google AI Studio, for closed-set decisions such as tool selection, policy checks, model routing, and context grading. It accepts one multimodal input (text, images, audio, video, PDFs up to 256K tokens) with up to 1,000 yes/no, multiple-choice, and rubric-scoring questions, scoring options in a single forward pass with confidence scores at 150-250 ms end-to-end latency and $0.10 per 1M input tokens with zero output tokens billed. Google reports that Gemini Decision 1 leads across agent routing, reasoning, ranking, and multimodal benchmarks with the lowest calibration error, and is nearly 5x cheaper than Gemini 3.5 Flash-Lite with the Interactions API across the shared benchmark suite.
Why it matters: It documents a new Google model and API with pricing, latency, and benchmark claims, useful for teams weighing low-latency classification and routing options.
Griyd is a multiplayer AI tool where teams plan together, write code with their own agents, and ship production-ready features, with a shared board showing who is connected and what each agent is building. It works with your own agents (Claude Code or Codex), your Git hosting, and your workflow, with CI continuing on GitHub or GitLab. Pricing is $15 per developer per month, self-hosting is $500 per year, members without agents are free, and there is a 14-day free trial with no credit card required.
A Show HN submission introduces AI Spar, a web app at aispar.app that lets Claude, GPT, and Gemini debate a question. The submission links to the app and to its Hacker News discussion thread.
Talos is an open-source, self-hosted AI agent for terminal and Telegram on macOS and Linux (including Raspberry Pi) that uses Ollama, the Claude Code CLI, or a configured API provider, and only executes tools after a deterministic permission kernel rules on each action, with authority bound to exact arguments as a single-use 30-second token. The 0.20.3 stable release adds native Telegram command blocks for known diagnostic arguments and separate measured result states, while retaining the frozen operator contract, permissions and dependency locks. The project states its limits plainly: shell execution requires bubblewrap or sandbox-exec and refuses to run unprotected otherwise, and it is not a multi-tenant security boundary.
Enspirit introduces Revu, a free, open-source, MIT-licensed tool (npm install -g @enspirit/revu) that runs locally on 127.0.0.1:7777 and provides a browser-based reading surface for reviewing agent-written code via "scopes" — sets of files sharing a property, configurable in a .revu.yml — rather than a file tree or diff. Remarks anchored to specific lines are tagged must-fix, question, or scope and exported as a single markdown document addressed to whoever writes the next change. The tool contains no AI by design and never writes to the project under review.
Mastra Forward is an annual service providing a named forward-deployed Mastra engineer who works in your codebase alongside your team, backed by Mastra core engineering. It includes Mastra Factory installed in your repo to run a swarm of agents, an Enterprise license on hosted, self-hosted, or BYOC deployment, a Trace Intelligence pool with evals run as a service, and a quarterly architecture review. Mastra says it costs less than one senior AI hire and is invoiced quarterly or prepaid at a discount.
As teams deploy and scale AI workloads using NVIDIA Dynamo, Kubeflow, or Slurm, tracking what actually runs requires visibility beyond the platform configuration captured in a recipe. Application teams define and update these workloads independently of the platform recipe, so a reproducible deployment bundle alone cannot establish which workload artifacts are running on each node.
Dedoomify is a tool that rewrites AI news stories to describe failures like ordinary software issues, for example turning "the model is misaligned" into "the model has a bug." Each change is highlighted with the original text available on hover or tap, and it runs on quick phrase rules or a local AI model in the browser via WebGPU. It is open source on GitHub.
agent-blast-radius is an Agent Skill that reads an AI agent's codebase and reports two things: its boundary (whether anything needs or reaches the outside world, such as hosted model APIs, telemetry, start-time package or model fetches, SaaS connectors and hard-coded hosts) and its blast radius (what the agent can do inside an enclave if steered by a poisoned document, an insider or a compromised host). It writes a BLAST_RADIUS.md report with file:line evidence, matched failure patterns, OWASP agentic (ASI) ids and the smallest fix per path, backed by a deterministic, read-only, offline Python scanner that never runs the target code or prints secret values. It installs as a Claude Code plugin (2.1.275+) or manually into any Agent Skills host, and is static-only, so network policy, IAM policies, token scopes and runtime MCP tools go under "Unverifiable statically."
RefTrail is a method and toolset that makes a coding agent's completion report checkable: each requirement gets a source and measurable acceptance criterion, the criteria are frozen with a SHA-256 digest before product code is written, and a coverage catalog records status, evidence and the limits of that evidence. Its tools require Node.js 22.16 or newer and no further packages, and a worked example rebuild of GNU uniq reports 100% functional coverage with 1/1 mandatory checks passed, with recorded output comparisons matching GNU uniq byte for byte (16 of 16 under Linux with version 9.4 for revision 1; for revision 2, 16 of 16 under Windows with version 8.32 and 16 of 16 under Linux with version 9.4). The project states that it organizes what you are allowed to observe rather than analyzing products, that the digest proves criteria did not change but not that they preceded the code, and that its legal notes are general information, not legal advice.
A Show HN post demonstrates using the FN2 skill to run a semantic search over 10,602,684 embedded passages from 142,758 earnings calls, answering a question about NVDA management commentary on Blackwell supply with quoted and cited results. The example shows a run completing in 35s using Qwen3-Embedding-8B (4096d), then scheduling a recurring pre-market NVDA brief with email delivery that resolves "before market open" against the exchange calendar. The post also shows an `agents` listing of existing scheduled agents, including one with 53 runs and two possible duplicates.
Google released A2UI 1.0, the first stable version of its open protocol for generative UI, which lets AI agents respond with interactive screens that each app renders with its own native components. Version 1.0 adds bidirectional function calls, accessibility support and extensions, plus open-source renderers for React, Angular, Lit, Flutter, SwiftUI and Jetpack Compose (currently v0.9.1) and agent SDKs for Python, TypeScript and Dart. Atlassian, Alation, Bloomberg and ByteDance are already building with it, and the web and Flutter renderers work inside MCP Apps such as Claude and Gemini.
Why it matters: It marks the first stable release of an open generative-UI protocol, with concrete renderer, SDK and partner details useful for developers evaluating agent interfaces.
Vapi has launched Agent Loop in early access, a self-improving harness on top of its voice agent platform that builds an agent from a plain prompt, scores every call against a business goal, clusters misses into issues, drafts and tests fixes, then ships them as experiments on a slice of live traffic with automatic rollback. It relies on primitives including Simulations and Traffic Splitting (available in the dashboard now) and Goals, Call Health and Experiments (coming soon), with human approval required before any change ships. GA is planned for early next year, and Vapi is taking waitlist signups ahead of VapiCon on November 11.
At its New York Summit, ElevenLabs announced new models, agents, and deployment options, including plans to fine-tune open-source language models and ElevenLabs voice models for businesses, a previewed Interaction Engine that manages turn-taking and selects LLMs per task, and ElevenAgents upgrades such as Memory and Routines. It also introduced ElevenAI, an app for enterprise teams that takes meeting notes and action items using Scribe v3, launched the ElevenLabs Partner Program and Marketplace, and announced an expanded NVIDIA partnership.
ElevenLabs launched the ElevenLabs Marketplace, a curated collection of industry-specific AI and voice applications built on ElevenLabs by partners including Salesforce, IBM, NiCE, Twilio and Zendesk, spanning financial services, insurance, healthcare, telecommunications, retail and government. Enterprises can explore solutions and connect directly with the partners behind them, while ElevenLabs says it plans to let customers purchase eligible partner solutions through existing ElevenLabs commitments starting next year.
This entry is a GitHub Changelog page for version 0.170.0-alpha.1, consisting mainly of site navigation menus covering GitHub products such as Copilot, Actions, Codespaces, and Advanced Security. The material contains no release notes or details about the version itself.
Sploom is a tool that lives alongside git and records which agent and model wrote each line of code, along with the intent, prompts, decisions and checks behind it, working with Claude Code, Codex, Cursor, Gemini CLI, Copilot CLI, Devin CLI, Grok Build, Goose, Cline and any MCP-capable agent. After a one-time `sploom init`, it installs git hooks so provenance travels with commits via git notes, and offers commands like `sploom why`, `sploom rewind --green` and `sploom context`. Prompts are encrypted to the team's SSH keys listed in a committed `.sploomkeys` file, and it runs locally with an initial ~130 MB embedding model download.
OpenShorts is an MIT-licensed, self-hostable tool that turns long videos into 9:16 shorts using Gemini for viral moment detection, faster-whisper subtitles, and AI reframing modes including face tracking and split-screen layouts. It also bundles an AI Shorts UGC creator and a YouTube Studio toolkit, and exposes the pipeline via an MCP server, REST API, webhooks and CLI. Self-hosting is free with bring-your-own API keys, while a hosted cloud plan starts at $12/mo after a watermarked 20 min/mo free tier.
Woolly is a Show HN project that provides a single local memory shared across all of a user's AI tools. The submission links to usewoolly.com and its Hacker News discussion thread.
Marktechpost launched GPU Leaderboard, a free cloud GPU price comparison site for AI inference that reads on-demand and spot prices from each provider's own pricing page, alongside a Chrome extension called GPU Leaderboard: Deploy It. The site ranks 201 live GPU offers from 24 providers by $/GPU-hr and includes a calculator that prices a full run including idle time, with an example 8×H100 72-hour run at 90% utilization ranging from $967.68 at Latitude to $3,545.28 at CoreWeave. The extension estimates VRAM needs and shows the cheapest fitting GPUs on Hugging Face model pages and GitHub repos, plus a toolbar lookup.
Google is introducing Tabletop, a new Google Labs experiment for real-time group collaboration where users add links, files and ideas, then prompt @tabletop or pick a starter template to generate interactive trip itineraries, dinner polls, expense splitters and tournament brackets. Once a space is ready, an invite link can be shared in a group chat so everyone can use it at the same time, like in a Google Doc. Tabletop is rolling out in the U.S. (18+) via waitlist at labs.google/tabletop.
Google Labs has introduced Tabletop, an experimental AI tool that generates customized collaborative hubs for coordinating with family and friends, with a built-in feature that lets users create personalized interactive tools like savings trackers, chore charts and event plans by describing them. Users can add links, files and ideas, then prompt @tabletop or pick a starter template to generate trip itineraries, dinner polls, expense splitters and tournament brackets, and share an invite link so others can use the space simultaneously like a Google Doc. The beta is rolling out via a waitlist and is currently limited to people over 18 in the US.
AWS announced usage-based metering for Amazon SageMaker AI bidirectional streaming inference, letting model providers report actual usage for AWS Marketplace software charges via a companion WebSocket metadata connection, while SageMaker AI endpoint infrastructure charges still apply. Deepgram is the launch partner, bringing usage-based Speech-to-Text and Text-to-Speech billing to AWS Marketplace. Model servers opt in with the X-Amzn-SageMaker-Metadata-Stream-Supported header and send JSON metering records with the inference.count dimension and ConsumedUnits.
AWS introduces the Universal Agent Evaluation Framework (UAEF), an open-source, framework-agnostic system that uses adapters to normalize agent traces from LangGraph, LangChain, Strands Agents, Amazon Bedrock AgentCore, and Langfuse into a common AgentTrace format, then scores them with 36 built-in metrics across 7 quality dimensions. Results are stored per experiment in Amazon DynamoDB and Amazon S3, enabling comparison of agent versions, regression detection, and report and insight generation, with both a Python library and a low-code web UI sharing the same engine. The post also provides a playbook for choosing metrics and preparing datasets, plus runnable notebooks for each supported framework.
Why it matters: It offers a detailed, practical reference for teams evaluating agents across frameworks, including metric selection guidance and a closed-loop workflow for catching regressions before release.
Databricks announced that Genie One is now the enterprise home for AI cowork, combining Genie Search (hybrid AI search indexing tools like Google Drive, Jira, Confluence, Workday and SharePoint, plus over 100 MCP connectors), organizational directories, in-conversation artifacts such as slide decks and dashboards, MCP-based actions across tools, shareable skills, and native apps for browser, desktop, mobile, Slack and Microsoft Teams. Genie can now take more autonomous action on users' behalf, with Unity Gateway providing traces, MCP policies, permissions and cost management, and Genie One is free for users through January 31, 2027.
Why it matters: It details a major enterprise AI assistant release, useful for readers tracking how agentic workplace tools are being packaged with governance and cross-tool search.