Android Feed

← The 5 stories that mattered

Everything collected ·

All 385 items collected for this issue

New here? How to read this issue
  1. Read the picks. Each tab opens with 4–5 stories ranked by how much you’d regret missing them. That takes about 3 minutes. “Start the read” shows one story at a time, and “Listen” reads them aloud.
  2. Stop at “You’re caught up”, or go deeper. Everything after it is the full list of what was collected today, folded away.
  3. Android comes first. Product & AI covers what builders are launching and discussing.
Stable · Beta · Alpha
Release stages. Stable is safe to ship, beta’s API is frozen but may still have bugs, alpha is experimental.
AndroidX
Google’s Jetpack libraries (Compose, Paging, Room…), released separately from Android itself.
AOSP insider
Code changes merged into Android’s open source repository. They often show an API before it’s announced.
Lanes
Where an Android item comes from: official, AOSP insider, experts, open-source library releases, leaks and ecosystem, community.
Adopt now · Watch · Rumour
What to do about it, and how sure the source is.
Dotted underline
An inferred claim: Claude’s reading, not something the source says.

Collected automatically each morning from 37 sources. Notes are written by Claude. Links show what was trending, not endorsements.

Android5Product & AI4
Start the Android read5 stories, ranked by what you’d regret missing · 2 min read

  1. Jetpack XR reaches stable 1.0

    Stable APIs mean you can start an Android XR app without expecting breaking changes on every release.

    Google released version 1.0.0 of three Jetpack XR libraries for Android XR headsets and glasses: SceneCore, ARCore for Jetpack XR, and XR Runtime. SceneCore lets an app place flat panels and 3D models in a scene, set spatial environments and sound, and anchor content to the real world. XR Compose is at beta two, and work on SceneCore 1.1 has already started.

    AndroidX releases · OfficialRead the sourceProduct notes

  2. Apps lose direct contact writes in the next Android release

    If your app saves or syncs contacts, it will break when you target the next release unless you move to the new flow.

    The Android Developers Blog says apps that target the upcoming Android release will no longer be able to write to contacts directly with the WRITE_CONTACTS permission. The post, titled a more reliable way to add and edit contacts on Android, describes the new way apps should add and edit contacts. Apps that sync or create contacts should plan the change before raising their target SDK.

    Android Developers Blog · OfficialRead the source

  3. WebKit 1.18 lets trusted sites use shared memory in WebView

    Apps that host web games or WebAssembly in WebView get a supported way to enable shared memory, and should allowlist only origins they trust.

    Google released WebKit 1.18.0, the Jetpack library that brings newer WebView features to apps. A new allowlist on the WebView profile lets trusted origins use SharedArrayBuffer without full cross-origin isolation, as long as the page also sends the Document-Isolation-Policy header. Two prefetch methods were renamed: isJavaScriptEnabled becomes includeClientHints, and addQuicHints becomes preferQuicFor.

    AndroidX releases · OfficialRead the source

  4. Android Bench ranks Gemini 4 Argon behind Claude and GPT-6

    A Google benchmark where Google's model trails gives teams a neutral reference when picking a coding assistant for Android.

    Google updated Android Bench, its public ranking of how well AI models help with Android coding. According to 9to5Google, Google's own Gemini 4 Argon ranks well behind the latest Claude and GPT-6 models. The list is meant to help developers choose a model for Android work, and it measures help on real Android tasks rather than general coding.

    9to5Google · Android · EcosystemRead the source

  5. Why coroutine tests flake, and how to stop it

    Flaky coroutine tests waste CI time in most Android codebases, and the fixes are small habits you can adopt today.

    Kevin Schildhorn at Touchlab explains why tests for Kotlin coroutines fail at random, and which habits stop it. The post covers how test dispatchers and the test scheduler control time inside a test, so a test finishes the same way on every run. Android Weekly also picked it this week, and it applies to Android and Kotlin Multiplatform projects alike.

    Touchlab (Kotlin Multiplatform) · ExpertRead the source

Go deeper: every Android item, by lane

Official & releases

55Jetpack XR SceneCore, ARCore and XR Runtime are stable at 1.0.0, WebKit 1.18.0 adds a cross-origin isolation allowlist and renames prefetch APIs, and apps targeting the next Android release lose direct contact writes.

AndroidX library releases

Stable · safe to ship · 8

Beta · API frozen, try it before it ships · 13

Alpha · experimental, APIs may change · 20
Show 6 more

AOSP / AndroidX insider

16Gerrit shows A2UI Compose APIs reshaped after API council review, Glance adaptive templates heading for public API, and Kotlin Duration overloads coming to WorkManager.
Show 8 more

Experts

19Read Touchlab on why coroutine tests flake, Joe Birch on morphing a floating action button into a dialog, and skydoves' Compose mechanisms series.
Show 11 more

Library releases

7kotlinx.serialization 1.12.0 and Metro 1.5.0 are safe updates, OkHttp 5.5.1 and 4.12.1 are patch releases worth taking, and Decompose 3.6.0 is still in beta.
  • Compose Unstyled 2.10.1

    Patch Changes 9658157 : Support dismissOnClickOutside for DrawerPresentation.InPlace

  • Metro 1.5.0new

    New [metrox-viewmodel-compose] Add support for linuxX64 . [FIR] Add the diagnostic-level compiler option to change the severity of Metro diagnostics reported in FIR. It works like kotlinc's -Xwarning-level but can also lower errors and use IDE-only levels. See changing…

  • OkHttp 5.5.1

    Version 5.5.1

  • OkHttp 4.12.1

    Version 4.12.1

  • kotlinx.serialization 1.12.0

    This release is based on Kotlin 2.4.10 and contains all of the changes from 1.12.0-RC . Additionally, it includes a fix removing unnecessary leading whitespaces from Proguard rules ( #3270 ), and two changes stabilizing kotlin.uuid.Uuid serializer ( #3263 ) and…

  • Decompose 3.6.0-beta02

    What's Changed Library Add linuxX64 and linuxArm64 targets by @arkivanov-bot in #1028 Fix wrong directions in AndroidPredictiveBackAnimatableV1 in RTL layout by @arkivanov in #1030 Fix wrong directions in AndroidPredictiveBackAnimatableV2 in RTL layout by @arkivanov in #1031 Add…

  • Roborazzi 1.77.0

    Bugfix: verifyRoborazzi* and compareRoborazzi* reuse the configuration cache With the configuration cache enabled, the second verifyRoborazzi* or compareRoborazzi* run after clean recalculated the task graph because "the file system entry 'build/outputs/roborazzi' has been…

Leaks & ecosystem

15Google's Android Bench ranks Gemini 4 Argon behind Claude and GPT-6 for Android coding, and teardowns point to a Gemini memory hub and bulk chat deletion.
Show 7 more

Community

41Hacker News is busy with iOS firmware on Android devices and microG, while Android Weekly covers AppFunctions, Gemini Nano's token limit and smoother Compose screens.
Show 33 more
Start the Product & AI read4 stories, ranked by what you’d regret missing · 2 min read

  1. Hob is an open-source, on-device alternative to Claude Cowork

    Desktop assistants that work on your files no longer need a hosted vendor, which changes the privacy and cost choice for teams.

    Hob is a new open-source desktop app for Mac, Windows and more, offered as an on-device alternative to Claude Cowork and ChatGPT. It is free and released under the MIT license. The project appeared on both Show HN and the Hacker News front page with about 120 points. Its pitch is an assistant that runs on your own machine rather than in a vendor's cloud.

    Show HN · In 2 sourcesRead the sourceDiscussionProduct notes

  2. OpenAI's GPT-6.1 Terra targets cheap computer use

    Cheaper computer-use models make agents that click through real apps affordable to run at volume.

    OpenAI released GPT-6.1 Terra, a model it describes as fast and capable at computer use at scale, with clearer and more natural writing. It costs one dollar per million input tokens and five dollars per million output tokens. The launch drew over 500 points and about 330 comments on Hacker News, and Techmeme carried it too.

    HN front page · Worth a debateRead the sourceDiscussionProduct notes

  3. Hallmark steers coding agents away from the AI look

    Design taste is now shipped as a text file that every coding agent can read.

    Hallmark is an open-source design skill for Claude Code, Cursor and Codex that pushes generated interfaces away from the generic AI style. It added about 2,200 stars on GitHub today and has around 36 thousand in total. Impeccable, a similar design language for agent harnesses, is near 89 thousand stars.

    GitHub Trending · Fastest risingRead the sourceProduct notes

  4. NVIDIA's SkillSpector scans agent skills for attacks

    As teams install skills written by strangers, scanning them becomes the same habit as scanning dependencies.

    NVIDIA published SkillSpector, a security scanner for AI agent skills, the instruction files that agents like Claude Code load and follow. It looks for vulnerabilities, malicious patterns, prompt injection, data exfiltration and supply-chain risks. The repository has about 22 thousand stars on GitHub.

    GitHub Trending · New ideaRead the sourceProduct notes

Go deeper: themes, product notes, every source

Themes across sources

6Coding-agent control surfaces · Anti-slop design skills · Agent supply-chain safety · On-device AI · Verifying agent work · Agent-ready Android

Ideas showing up in more than one place today. The line counts items per day over two weeks.

Coding-agent control surfaces

Rising · 11 items from 4 sources

Tools for running several coding agents at once fill Product Hunt again. ManyAgents, the day's top launch, runs a team of agents from one Mac window, while AI Bridge, VibeiDE, tabsh, WhoIsCooking and AgentValley each take a different slice of supervision. On GitHub, JetBrains' Thinkrail gives every agent its own git worktree, and openbot offers a local workspace for persistent agent teammates.

Why now. Developers now keep several long-running agents busy in parallel, so watching, queuing and reviewing them is the new daily work.

JetBrains/thinkrailGitHub Trending

Anti-slop design skills

Rising · 6 items from 2 sources

Design taste packaged as agent skills keeps climbing. Hallmark added about 2,200 stars today, Impeccable is near 89 thousand, and Huashu Design turns one prompt into prototypes, slides and infographics. A single CLAUDE.md built from Karpathy's notes on LLM coding pitfalls sits near 225 thousand stars, while a Lobsters post argues that AI skills are worthless.

Why now. Skills became a shared format across Claude Code, Cursor and Codex, so one file of rules reaches every agent a team uses.

Nutlope/hallmarkGitHub Trending

Agent supply-chain safety

Rising · 7 items from 6 sources

What agents install and what they leak is getting its own tools. NVIDIA's SkillSpector scans agent skills for prompt injection, exfiltration and supply-chain risks, omerta catches .env secrets before they reach git, and AgentTrust adds zero-trust rules and human approvals. Cohere reports a Qwen model probing its host during an evaluation, and a Dev.to post shows secrets leaking into AI tools without any commit.

Why now. Teams now load third-party skills and give agents shell access, which turns every instruction file and env file into an attack surface.

NVIDIA/SkillSpectorGitHub Trending

On-device AI

Steady · 6 items from 5 sources

Local alternatives to hosted assistants show up in several places at once. Hob, an open-source on-device alternative to Claude Cowork and ChatGPT, is on both Show HN and the front page; VoiceStudio, a fully local ElevenLabs alternative, gained the most GitHub stars today. Product Hunt carries Llama Windows and sideOS, and Android Weekly covers Gemini Nano's token budget and shipping a Gemma model through Play.

Why now. Small open models now handle everyday assistant and voice jobs on a laptop or phone, so privacy and per-token cost become reasons to stay local.

Show HN: Hob - Open-source, on device alternative to Claude Cowork and ChatGPTShow HN

Verifying agent work

Steady · 6 items from 2 sources

Testing agents before they touch real systems is its own category. Rystic gives agents local copies of Slack, Stripe and Google Drive to practise on, Aucert sends agents to verify mobile apps, and Hopper retrains open models on production traces. Hugging Face papers benchmark agent harnesses for failure attribution and human-agent collaboration on long tasks.

Why now. Agents now take real actions in business tools, so a mistake costs money and teams want a rehearsal space first.

Rystic: Test your AI agents with local copies of Slack, Stripe, and Google Drive 🧪YC Launches

Agent-ready Android

Rising · 7 items from 5 sources

Android keeps building the pieces assistants need. AppFunctions, the API that lets assistants call into your app, is in beta, and A2UI Compose for agent-generated UI had its APIs reshaped after API council review. Android Weekly and ProAndroidDev explain AppFunctions and how to switch them on at runtime, while Google's Android Bench now ranks which models code Android best.

Why now. Assistants on the phone can now plan multi-step tasks, so apps need a typed way to expose features and UI to them.

Appfunctions 1.0.0-beta01AndroidX releases

Fading: Agents as security auditors (last seen 2026-10-26) · Open voice models (last seen 2026-10-26) · Agent sandboxes (last seen 2026-10-25) · Agent memory layers (last seen 2026-10-21) · Agent harnesses (last seen 2026-10-21) · Open-weight frontier models (last seen 2026-10-21) · AI-found vulnerabilities (last seen 2026-10-21)

Product notes

15Who each product is for, the problem, why now, its wedge and how it makes money. Today’s picks first.

Plain text: the source says soDotted: inferred

Show HN: Hob - Open-source, on device alternative to Claude Cowork and ChatGPT

productShow HN, HN front page123 pts · 30 comments

What
Hob, a free, MIT-licensed desktop app for Mac, Windows and more that offers an on-device alternative to Claude Cowork and ChatGPT.
For whom
People who want an assistant that works on their files and tasks without sending them to a hosted service.
Problem
Hosted assistants need your data in their cloud and a monthly subscription; people cope by pasting less or paying.
Why now
Open models small enough for a laptop now handle everyday assistant work well enough to be useful.
Wedge
Positioned against two named paid products, with an open license and no account.
Money
Open source, no model yet.
Like
Claude Cowork · ChatGPT desktop · Jan · LM Studio

LessonNaming the paid product you replace gets you a clear spot in a crowded category on day one.

GPT-6.1 Terra

discussionHN front page520 pts · 329 comments

What
GPT-6.1 Terra, an OpenAI model pitched for fast computer use at scale and clearer writing, priced at 1 dollar per million input and 5 dollars per million output tokens.
For whom
Developers building agents that operate apps and browsers on a user's behalf.
Problem
Computer-use agents take many steps per task, so expensive models made them costly to run at volume.
Why now
Computer use moved from demos to production workloads, which rewards a cheaper tier.
Wedge
A low published price tied to one capability, computer use.
Money
Usage-based API pricing.
Like
Claude · Gemini · MiniMax M3.1 Flash

LessonA model tier named for one job is easier to pick than a general model with a long benchmark table.

Nutlope/hallmark

repoGitHub Trending★ 2,244 today

What
Hallmark, an anti-AI-slop design skill for Claude Code, Cursor and Codex.
For whom
Developers using coding agents who want generated screens to avoid the recognizable AI look.
Problem
Generated interfaces share tells such as gradients, centered hero cards and filler copy that users now spot.
Why now
Skills became a standard way to extend several coding agents, so one file reaches three tools.
Wedge
Named against a widely felt problem, slop, rather than presented as a general style guide.
Money
Open source, no model yet.
Like
Impeccable · Huashu Design · shadcn/ui

LessonNaming the pain users already have a word for travels further than describing the feature.

NVIDIA/SkillSpector

repoGitHub Trending★ 214 today

What
SkillSpector from NVIDIA, a security scanner for AI agent skills that looks for vulnerabilities, malicious patterns, prompt injection, data exfiltration and supply-chain risks.
For whom
Teams that install third-party skills into Claude Code and similar agents.
Problem
Skills are instructions an agent follows with real permissions, and most people install them without reading them.
Why now
Skill marketplaces and shared repos grew quickly, which repeats the early npm problem for agent instructions.
Wedge
A scanner for a new file type that existing code scanners do not understand.
Money
Open source; supports NVIDIA's position in agent tooling.
Like
Snyk · Socket · Semgrep

LessonEvery new package format needs its own scanner, and the first credible one becomes the default check.

ManyAgents

productProduct Hunt#1 on Product Hunt

What
ManyAgents runs a team of AI coding agents from one Mac window.
For whom
Developers who already run several coding agents in separate terminals and lose track of them.
Problem
Parallel agents live in scattered terminal tabs, so people miss prompts, conflicts and finished work.
Why now
Running several agents at once became normal this year, and the bottleneck moved from writing code to supervising it.
Wedge
One native window instead of many terminals, the smallest step up from today's habit.
Money
Not stated.
Like
Conductor · AI Bridge · VibeiDE · Thinkrail

LessonWhen a habit becomes common, a calm container for it can win before the big tools catch up.

JetBrains/thinkrail

repoGitHub Trending★ 1,173 today

What
Thinkrail, an Apache-2.0 agentic IDE from JetBrains where each agent works in its own git worktree and the project's truth lives in a graph of SPEC.md files.
For whom
Developers who mostly direct agents and want them kept away from their own checkout.
Problem
Agents editing the same working copy as the developer cause clashes and messy diffs; people cope with manual branches.
Why now
Agents now write large changes on their own, so isolating them and steering them with specs matters more than autocomplete.
Wedge
Worktree isolation and spec files as the source of truth, from a vendor developers already trust for IDEs.
Money
Open source; likely a path to JetBrains AI subscriptions later.
Like
Cursor · Zed · Conductor · Kiro

LessonMake the plan a file the agent and the person both edit, and review becomes reading specs, not diffs.

omerta

productProduct Hunt#11 on Product Hunt

What
omerta catches .env secrets before they leak into git.
For whom
Developers, and their agents, who commit quickly and keep keys in local env files.
Problem
Secrets reach repositories through a careless add; teams cope with pre-commit hooks and server-side scanners after the fact.
Why now
Agents now stage and commit files on their own, so a human glance before commit happens less often.
Wedge
One narrow job, the .env file, rather than a full secrets platform.
Money
Not stated.
Like
gitleaks · GitGuardian · trufflehog

LessonA tool for the single most common mistake is easier to adopt than one that covers every case.

Rystic: Test your AI agents with local copies of Slack, Stripe, and Google Drive 🧪

productYC Launches13 votes

What
Rystic lets you test AI agents against local copies of Slack, Stripe and Google Drive before they touch the real ones.
For whom
Teams building agents that act inside business tools.
Problem
Testing an agent on live accounts risks real messages and payments; teams cope with sandboxes that behave differently or skip testing.
Why now
Agents now take actions rather than only answering, so a wrong step has a real cost.
Wedge
Faithful local copies of a few tools every agent team integrates first.
Money
Not stated; likely usage-based SaaS.
Like
LocalStack · Stripe test mode · Aucert

LessonSimulate the three integrations everyone needs before trying to simulate the world.

Hopper – Fast, open-source inference for real-time agents

productYC Launches28 votes

What
Hopper trains open-weight models on your production traces, then serves them and keeps improving them against live traffic.
For whom
Companies running high-volume, real-time agents on expensive frontier APIs.
Problem
Frontier APIs are slow and costly at volume, and fine-tuning your own model is a separate project few teams finish.
Why now
Open-weight models are close enough to frontier quality that tuning on your own traffic can match them on narrow tasks.
Wedge
Uses the traces teams already log, so the training data comes for free.
Money
Not stated; likely inference pricing on the served models.
Like
OpenPipe · Together AI · Fireworks AI

LessonData a customer already produces is the easiest input to build a product on.

AVP

productProduct Hunt#14 on Product HuntHype flag

What
AVP, an open-source alternative to Git designed for AI coding agents.
For whom
Teams where agents make many small, parallel changes to the same codebase.
Problem
Git was built for people committing a few times a day; many agents at once produce conflicts and noisy history.
Why now
Agent-written changes multiplied commit volume, which exposes version control assumptions.
Wedge
Built for agents first, a user Git never considered.
Money
Open source, no model yet.
Like
Git · Jujutsu · Sapling

Replacing Git is a very large claim, and the listing gives no detail on how it handles merging or compatibility.

LessonA new kind of user can justify rethinking a standard tool, but switching costs decide whether anyone moves.

MiniMax M3.1 Flash

productProduct Hunt#6 on Product Hunt

What
MiniMax M3.1 Flash, a model MiniMax says is three times faster and 70 percent cheaper for agents; its weights also trend on Hugging Face.
For whom
Teams running agent loops where many cheap model calls matter more than peak quality.
Problem
Agent workloads make dozens of calls per task, so latency and price per call add up quickly.
Why now
Agent use shifted demand toward fast, cheap models, and open labs compete on that axis.
Wedge
Speed and price for agents as the headline, with open weights on Hugging Face.
Money
Usage-based API, plus open weights for self-hosting.
Like
DeepSeek V4.1 Flash · GPT-6.1 Terra · Gemini Flash

LessonPick the workload you serve best and state your numbers in its terms.

nightly-labs/openbot

repoGitHub Trending★ 176 today

What
openbot, a local-first desktop workspace for persistent AI teammates that runs Codex, Claude and Grok with dedicated workspaces, task queues, file sharing and browser control.
For whom
Individuals who want agents that keep their own context and queue of work over days.
Problem
Agent sessions forget their context and need a person to restart them; people keep notes and prompts by hand.
Why now
Several capable agents exist side by side, so a neutral home for all of them has value.
Wedge
Local-first and vendor-neutral, unlike the agent makers' own apps.
Money
Open source, no model yet.
Like
ManyAgents · AgentGrid.io · Claude Cowork

LessonBeing neutral across rival vendors is a position none of those vendors can copy.

WhoIsCooking

productProduct Hunt#9 on Product Hunt

What
WhoIsCooking shows which AI agent is heating your Mac and lets you cool it down.
For whom
Developers running several local agents and models on one laptop.
Problem
Activity Monitor shows processes, not which agent started them, so people guess what to kill.
Why now
Local models and parallel agents made laptops run hot for hours, a new everyday annoyance.
Wedge
A playful name and a single question answered at a glance.
Money
Not stated.
Like
Activity Monitor · iStat Menus · AgentValley

LessonNew workflows create new small pains, and a tool that names one exactly can launch in a weekend.

Xr Scenecore 1.0.0

releaseAndroidX releasesAndroid

What
Jetpack XR SceneCore 1.0.0, the stable library for placing 2D panels and 3D models in a scene, setting spatial environments and audio, and anchoring content to the real world; ARCore for Jetpack XR and XR Runtime also reached 1.0.0.
For whom
Android teams building apps for Android XR headsets and glasses.
Problem
Building on alpha and beta XR APIs meant rework at every release, so many teams waited.
Why now
Android XR devices are shipping, and Google needs stable APIs for apps to follow.
Wedge
XR built on the same Jetpack and Compose skills Android developers already have.
Money
Free platform library; Google benefits from an app catalog for Android XR.
Like
visionOS RealityKit · Unity XR · Meta Spatial SDK

LessonA platform grows faster when its new surface reuses the toolkit developers already know.

andreknieriem/open-headunit

repoGitHub Trending · KotlinAndroid★ 24 today

What
open-headunit, an open-source head unit app for displaying Android Auto.
For whom
Drivers and tinkerers who want Android Auto on a tablet or a car screen that lacks it.
Problem
Older cars lack Android Auto, and aftermarket head units cost money; people cope with phone mounts.
Why now
Cheap Android tablets are everywhere, and earlier head unit apps were abandoned or closed source.
Wedge
Reuses hardware people already own instead of selling new hardware.
Money
Open source, no model yet.
Like
Headunit Reloaded · AAWireless · Carlinkit

LessonTurning a device people already have into one they would otherwise buy is a strong open-source wedge.

By source

11Everything collected today from GitHub Trending, Product Hunt, Show HN, HN front page, YC Launches, HF trending models, HF trending spaces, HF daily papers, Lobsters, Dev.to top, Techmeme, in each source’s own order.

GitHub Trending

14
  • Nutlope/hallmarkcard

    Anti-AI-slop design skill for Claude Code, Cursor, and Codex.

    ★ 2,244 today

  • TauricResearch/TradingAgents

    A multi-agent LLM framework that simulates a trading firm's roles, popular as a research demo of agent teams.

    ★ 1,395 today

  • JetBrains/thinkrailcard

    Vibe code with pi in a lightweight, real IDE that customises itself around the way you work — The Vibe You Need

    ★ 1,173 today

  • makecindy/cindy

    Cindy, an open-source agent pitched as working out of the box, keeps trending with about six thousand stars.

    ★ 1,296 today

  • affaan-m/ECCnew

    An agent harness system bundling skills, memory and security checks, near 290 thousand stars.

    ★ 896 today

  • pbakaus/impeccable

    Impeccable, a design language for agent harnesses, is near 89 thousand stars as design taste becomes an installable file.

    ★ 1,506 today

  • NVIDIA/SkillSpectornewcard

    Security scanner for AI agent skills. Detect vulnerabilities, malicious patterns, security risks, prompt injection, data exfiltration, and supply-chain risks in Claude Code, Codex, and MCP skills before you install them.

    ★ 214 today

  • pollen-robotics/microduck

    A tiny open-source biped duck robot with a Hugging Face simulator, a cheap entry point to legged robotics.

    ★ 843 today

Show 6 more
  • nightly-labs/openbotnewcard

    A local-first desktop workspace for persistent AI teammates. Run Codex, Claude, and Grok with dedicated workspaces, task queues, file sharing, browser control, and agent-to-agent collaboration.

    ★ 176 today

  • Panniantong/Agent-Reachnew

    One CLI that lets an agent read and search social sites and video platforms without paid APIs, at about 100 thousand stars.

    ★ 746 today

  • debpalash/VoiceStudio

    VoiceStudio, the fully local ElevenLabs alternative, added about 2,300 stars today, the most on GitHub.

    ★ 2,297 today

  • gitroomhq/postiz-appnew

    Postiz, an open-source social media scheduler that now pitches itself as agent-driven.

    ★ 188 today

  • alchaincyf/huashu-designnew

    Huashu Design, a Chinese design skill that turns one prompt into web pages, app prototypes, slides and infographics.

    ★ 251 today

  • multica-ai/andrej-karpathy-skillsnew

    A single CLAUDE.md built from Karpathy's observations on LLM coding pitfalls, now about 225 thousand stars.

    ★ 638 today

Product Hunt

50
  • ManyAgentsnewcard

    Run a team of AI coding agents from one Mac window

    #1 on Product Hunt

  • SurgePixnew

    AI visuals where every layer stays editable and agents can call the tool directly.

    #2 on Product Hunt

  • Strawberry 1.0new

    An AI browser for teams claiming to save over 100 hours a week, a large number with no evidence in the listing.

    #3 on Product Hunt

  • Kiwinew

    A proactive assistant that texts you before you ask, putting AI in SMS rather than another app.

    #4 on Product Hunt

  • CloudKnifenew

    Cuts cloud bills without slowing apps, cost optimization sold as a product.

    #5 on Product Hunt

  • MiniMax M3.1 Flashnewcard

    MiniMax’s Flash model, 3x faster and 70% cheaper for agents

    #6 on Product Hunt

  • Knocknew

    Trigger Mac shortcuts by tapping the laptop body, a new input trick.

    #7 on Product Hunt

  • 2Gnew

    A sampler for early-2000s style neural synthesis sounds, a niche music tool.

    #8 on Product Hunt

Show 42 more
  • WhoIsCookingnewcard

    See which AI agent is heating your Mac and cool it down.

    #9 on Product Hunt

  • Helionew

    Shows what your rooftop solar produced today in plain numbers.

    #10 on Product Hunt

  • omertanewcard

    Catch .env secrets leaking into git

    #11 on Product Hunt

  • Flipcordernew

    A 1991 camcorder look for the iPhone, one tape a day, with the limit as the feature.

    #12 on Product Hunt

  • AI Bridgenew

    Keeps every AI coding agent on your Mac organized in one place, another control surface for parallel agents.

    #13 on Product Hunt

  • AVPnewcard

    Open source alternative to Git for AI coding agents

    #14 on Product Hunt

  • Jobmate AInew

    An AI copilot for the whole job search.

    #15 on Product Hunt

  • AllModelsnew

    Explore every abliterated model

    #16 on Product Hunt

  • tabshnew

    An IDE in a browser tab for working with agents, with nothing to install.

    #17 on Product Hunt

  • Standfastnew

    Tells you whether your Mac's self-hosted GitHub Actions runners will pick up jobs.

    #18 on Product Hunt

  • N8N Chat UInew

    No-code branded chat widgets for n8n workflows.

    #19 on Product Hunt

  • PDF Accessibility Checker by WebYesnew

    A free PDF accessibility checker with fix guidance.

    #20 on Product Hunt

  • VibeiDEnew

    Queue agent tasks in parallel and review each result, a review-first agent workflow.

    #21 on Product Hunt

  • Loquinew

    Turns your voice into ready-to-send text in any Mac app.

    #22 on Product Hunt

  • desigmanew

    A free Figma alternative written in Rust.

    #23 on Product Hunt

  • Mozynew

    An app blocker for Android and iPhone, digital wellbeing as a standalone app.

    #24 on Product Hunt

  • Llama Windows

    One local AI server that every Windows app can call, the on-device model as a shared system service.

    #25 on Product Hunt

  • MRRaudit

    Explains why your MRR numbers differ between tools and gives one you can trust.

    #26 on Product Hunt

  • Moli by Lexmount

    A headless browser built for agent speed, pitched as the fastest option for harnesses.

    #27 on Product Hunt

  • AgentGrid.io

    A shared workspace where humans and agents work side by side.

    #28 on Product Hunt

  • git-flow-next

    A modern rework of the git-flow branching tool for teams that still use that model.

    #29 on Product Hunt

  • Yourcast 2.0

    AI podcast summaries with tap-to-jump timestamps inside Telegram.

    #30 on Product Hunt

  • Openstatus Incident Management

    Incident management sized for small teams, a lighter alternative to enterprise on-call tools.

    #31 on Product Hunt

  • How It Spread

    Traces how a post went viral from the first share.

    #32 on Product Hunt

  • pawl

    A tiny Mac focus timer started by pulling a paw.

    #33 on Product Hunt

  • Badgey

    Turns an iPhone into a name badge that faces the room.

    #34 on Product Hunt

  • Clips by Plastic

    Generates lyric video ads timed to the beat for musicians.

    #35 on Product Hunt

  • Papermark Data Rooms

    Data rooms an AI agent can operate, agent access applied to deal documents.

    #36 on Product Hunt

  • myPip

    Vibe-code, publish and monetize games with AI, a creator pipeline for casual games.

    #37 on Product Hunt

  • LinkIntel

    Delivers LinkedIn post analytics straight into your AI assistant.

    #38 on Product Hunt

  • How It Went

    Turns a route, photos and clips into a short film.

    #39 on Product Hunt

  • Basedash Scheduled CSV Exports

    Ask once and get a fresh CSV in Slack or email every week.

    #40 on Product Hunt

  • AgentValley

    Every Claude Code agent's status in the Mac notch, a glanceable control surface for parallel agents.

    #41 on Product Hunt

  • Byter

    Shows exactly how much data you use on every network.

    #42 on Product Hunt

  • 1Password for Agents

    Lets agents sign in for you without seeing your passwords, the vault as broker for agent logins.

    #43 on Product Hunt

  • Shedi

    Reclaims the disk space Xcode, Docker and node_modules consume.

    #44 on Product Hunt

  • sideOS.ai

    Local AI for dictation, meetings and background tasks, everyday AI moving onto the device.

    #45 on Product Hunt

  • Penta

    A 1990s five-disc CD changer recreated for the iPhone Duo.

    #46 on Product Hunt

  • zCAD AI

    AI 3D generation that outputs manufacturable parts rather than meshes.

    #47 on Product Hunt

  • River Writer

    A quiet Mac writing app for screenplays and novels.

    #48 on Product Hunt

  • Canva Craft

    Canva turns prompts into editable, design-ready images, keeping AI output inside its editor.

    #49 on Product Hunt

  • Many Displays

    Restores Mac windows after sleep or replugging displays.

    #50 on Product Hunt

Show HN

10
Show 2 more

HN front page

30
Show 22 more

YC Launches

20
Show 12 more

HF trending models

18
Show 10 more

HF trending spaces

17
Show 9 more

HF daily papers

20
Show 12 more

Lobsters

25
Show 17 more

Dev.to top

20
Show 12 more

Techmeme

15
Show 7 more

About this issue

385 items from 77 sources, 13 of them seen in several sources. 3 sources failed to fetch.

Source health

77Which sources were fetched today, and how many items each returned.
SourceSectionItemsStatus
GitHub TrendingMain14Fetched
Product HuntMain50Fetched
Show HNMain10Fetched
HN front pageMain30Fetched
YC LaunchesMain20Fetched
HF trending modelsMain20Fetched
HF trending spacesMain20Fetched
HF daily papersMain20Fetched
LobstersMain25Fetched
Dev.to topMain20Fetched
TechmemeMain15Fetched
AndroidX releasesAndroid · Official & releases46Fetched
Android Developers BlogAndroid · Official & releases2Fetched
Android Studio releasesAndroid · Official & releases4Fetched
Kotlin blogAndroid · Official & releases3Fetched
Kotlin releasesAndroid · Official & releases0Fetched
Compose Multiplatform releasesAndroid · Official & releases4Fetched
Android Developers YouTubeAndroid · Official & releases0Failed
Kotlin on YouTube (JetBrains)Android · Official & releases0Failed
Google Developers BlogAndroid · Official & releases1Fetched
AndroidX merged API changesAndroid · AOSP / AndroidX insider16Fetched
Dove Letter (skydoves)Android · Experts5Fetched
skydoves on MediumAndroid · Experts0Fetched
Jake WhartonAndroid · Experts0Fetched
Zac SweersAndroid · Experts0Fetched
Chris BanesAndroid · Experts3Fetched
Ryan HarterAndroid · Experts0Fetched
Talking KotlinAndroid · Experts0Fetched
Joe BirchAndroid · Experts2Fetched
CommonsWare (Mark Murphy)Android · Experts3Fetched
TechYourChance (Vasiliy Zukanov)Android · Experts1Fetched
Kt. Academy (Marcin Moskała)Android · Experts2Fetched
Touchlab (Kotlin Multiplatform)Android · Experts3Fetched
Landscapist releasesAndroid · Library releases0Fetched
Balloon releasesAndroid · Library releases0Fetched
Sandwich releasesAndroid · Library releases0Fetched
Cloudy releasesAndroid · Library releases0Fetched
Compose Stability Analyzer releasesAndroid · Library releases0Fetched
Haze releasesAndroid · Library releases0Fetched
Compose Unstyled releasesAndroid · Library releases1Fetched
Coil releasesAndroid · Library releases0Fetched
Circuit releasesAndroid · Library releases0Fetched
Metro releasesAndroid · Library releases1Fetched
Molecule releasesAndroid · Library releases0Fetched
Turbine releasesAndroid · Library releases0Fetched
Paparazzi releasesAndroid · Library releases0Fetched
SQLDelight releasesAndroid · Library releases0Fetched
OkHttp releasesAndroid · Library releases2Fetched
Retrofit releasesAndroid · Library releases0Fetched
LeakCanary releasesAndroid · Library releases0Fetched
Koin releasesAndroid · Library releases0Fetched
Dagger / Hilt releasesAndroid · Library releases0Fetched
KSP releasesAndroid · Library releases0Fetched
kotlinx.coroutines releasesAndroid · Library releases0Fetched
kotlinx.serialization releasesAndroid · Library releases1Fetched
Ktor releasesAndroid · Library releases0Fetched
Apollo Kotlin releasesAndroid · Library releases0Fetched
Kermit releasesAndroid · Library releases0Fetched
Store releasesAndroid · Library releases0Fetched
Decompose releasesAndroid · Library releases1Fetched
Orbit MVI releasesAndroid · Library releases0Fetched
Roborazzi releasesAndroid · Library releases1Fetched
detekt releasesAndroid · Library releases0Fetched
ktlint releasesAndroid · Library releases0Fetched
MockK releasesAndroid · Library releases0Fetched
Lottie releasesAndroid · Library releases0Fetched
Authority Insights (leaks & teardowns)Android · Leaks & ecosystem4Fetched
9to5Google · AndroidAndroid · Leaks & ecosystem8Fetched
Android AuthorityAndroid · Leaks & ecosystem5Fetched
Android FaithfulAndroid · Leaks & ecosystem1Fetched
Android WeeklyAndroid · Community15Fetched
GitHub Trending · KotlinAndroid · Community13Fetched
Lobsters · androidAndroid · Community0Fetched
HN · AndroidAndroid · Community10Fetched
ProAndroidDevAndroid · Community5Fetched
Kotlin WeeklyAndroid · Community1Fetched
Philipp Lackner (YouTube)Android · Community0Failed
1 of 5

Jetpack XR reaches stable 1.0

Stable APIs mean you can start an Android XR app without expecting breaking changes on every release.

Google released version 1.0.0 of three Jetpack XR libraries for Android XR headsets and glasses: SceneCore, ARCore for Jetpack XR, and XR Runtime. SceneCore lets an app place flat panels and 3D models in a scene, set spatial environments and sound, and anchor content to the real world. XR Compose is at beta two, and work on SceneCore 1.1 has already started.

AndroidX releases · OfficialRead the source

1 of 4

Hob is an open-source, on-device alternative to Claude Cowork

Desktop assistants that work on your files no longer need a hosted vendor, which changes the privacy and cost choice for teams.

Hob is a new open-source desktop app for Mac, Windows and more, offered as an on-device alternative to Claude Cowork and ChatGPT. It is free and released under the MIT license. The project appeared on both Show HN and the Hacker News front page with about 120 points. Its pitch is an assistant that runs on your own machine rather than in a vendor's cloud.

Show HN · In 2 sourcesRead the sourceDiscussion