Android Feed

← The 5 stories that mattered

Everything collected ·

All 366 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. AppFunctions lets your app offer actions to on-device AI

    This is how your app's features could become something an assistant can do for users without opening your app.

    A new explainer walks through AppFunctions, an Android platform API with a Jetpack library. You mark suspend functions with an annotation, and agents such as Gemini can discover and call them on the device, much like tools on an MCP server. It needs Android 16, it is still an experimental preview, and Gemini access is limited to testers for now.

    ProAndroidDev · CommunityRead the source

  2. kotlinx.serialization 1.12.0 is stable

    It is a low-risk update for any Kotlin app already on Kotlin 2.4.

    JetBrains released version 1.12.0 of kotlinx.serialization, the official Kotlin library for turning objects into JSON and other formats. It is built on Kotlin 2.4.10, includes everything from the release candidate, stabilises the serializer for Kotlin's Uuid type and removes stray whitespace from its Proguard rules.

    kotlinx.serialization releases · LibraryRead the source

  3. Media3 1.12 begins its alpha

    Video apps can start trying frame-level metadata and wide-colour features before they reach stable.

    Media3, Android's Jetpack library for audio and video playback, released 1.12.0 alpha01. It adds a constant for the Display P3 colour space, a shared registry for feature flags, playlist and category fields in media metadata, and listeners that let ExoPlayer apps read metadata attached to each video frame.

    AndroidX releases · OfficialRead the source

  4. Android 17 QPR3 Beta 2 reaches Pixels with 21 fixes

    If your app shows notification counts or targets Pixels, test it on the new beta.

    Google released the second beta of Android 17 QPR3, the quarterly platform update due in March 2027, for Pixel phones. According to 9to5Google it brings 21 fixes. Separate code findings suggest Pixel launcher will get numeric unread badges on app icons, a feature other Android skins have had for years.

    9to5Google · Android · EcosystemRead the source

  5. Not every coding-agent task needs a plan

    It is practical advice on when planning saves time with agents and when it only slows you down.

    Chris Banes, maintainer of the Haze blur library for Compose, compares two real tasks done with a coding agent. One went fine with direct implementation, the other needed a written plan first. He explains how to tell the two apart and which checks matter either way, such as tests and reviewing the diff.

    Chris Banes · ExpertRead the source

Also worth knowing

Releases, policy and leaks an Android engineer should know about this week.

Go deeper: every Android item, by lane

Official & releases

29kotlinx.serialization 1.12.0 is stable and safe to adopt on Kotlin 2.4, Media3 1.12 starts its alpha with per-frame video metadata, and Android Studio Rabbit 2 reached release candidate.

AndroidX library releases

Stable · safe to ship · 2

Beta · API frozen, try it before it ships · 1

Alpha · experimental, APIs may change · 3
Show 15 more

AOSP / AndroidX insider

24Remote Compose keeps gaining its own drawing and layout types, WebView in Compose is getting a proper bridge, and a new people client library points to a public people API.
  • Compose: MutableVector deprecatednew

    Deprecate MutableVector

    Compose is moving off its own vector type toward standard or newer collection types, likely from the AndroidX Collection library.

  • New people client librarynew

    Add androidx.people:people-client

    A people-client artifact suggests a public API for reading people and interaction data, following last week's people directory provider.

  • Compose: simple remember factories deprecatednew

    Deprecate remember factories that only wrap a constructor

    Helpers that only wrap a constructor are being removed in favour of calling remember with the constructor directly, trimming API surface.

  • Internal flag for process isolationnew

    Add internal process isolation flag and bridge

    A bridge for running parts of a library in an isolated process points to sandboxing, possibly for WebView or AI features.

  • Text layout: boxes and paths for a rangenew

    Add getBoundingBoxesForRange and fillPathForRange to TextLayoutResult

    Getting bounding boxes and fill paths for a text range makes custom highlights and selection effects easier in Compose.

  • Remote Compose preview tidy-upnew

    Remove redundant remoteDensity assignment in RemoteTextFontScaleComparisonPreview

    Small cleanup in Remote Compose previews; no API change.

  • Stats logger renamednew

    Rename ProviderStatsLogger to PlStatsLogger

    An internal rename with no user-facing effect.

  • WebView in Compose: new Chrome client bridgenew

    web-compose: Add WebChromeClientBridge and deprecate fullscreen View APIs

    A bridge for WebChromeClient and deprecated fullscreen View APIs show WebView support in Compose maturing into a proper API.

Show 16 more

Experts

16Chris Banes shows when a coding agent needs a plan and when to let it implement directly, and Dove Letter's part two explains how the Compose runtime remembers state.
Show 8 more

Library releases

3kotlinx.serialization 1.12.0 is the release worth updating for; Compose Unstyled 2.10.1 is a small patch and Decompose 3.6.0 is still beta.
  • Compose Unstyled 2.10.1

    Patch Changes 9658157 : Support dismissOnClickOutside for DrawerPresentation.InPlace

  • 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…

Leaks & ecosystem

21Android 17 QPR3 Beta 2 brings 21 fixes for Pixel testers, and leaks point to numeric notification badges on the Pixel launcher and work-profile apps in Android Auto.
Show 13 more

Community

40Read the AppFunctions explainer on making your app callable by on-device AI; an open-source AirPlay receiver for Android and a microG thread also drew attention.
Show 32 more
Start the Product & AI read4 stories, ranked by what you’d regret missing · 1 min read

  1. AI finds two critical PHP flaws that affect much of the web

    AI-found bugs in core infrastructure are arriving faster, and patching windows are getting shorter.

    PHP's maintainers patched two critical flaws, nicknamed PHP2Shell, that let attackers run code on servers without logging in. OpenAI says its GPT-5.6 Sol model found them in August. Fixes cover only PHP 8.2 and newer, Cloudflare added firewall protection before disclosure, and some European public websites went offline while teams patched.

    HN front page · Across sourcesRead the sourceDiscussion

  2. Cloudflare releases a security audit skill for coding agents

    Teams can now point their own agent at their code for a structured security review.

    Cloudflare published a skill that teaches coding agents to run a security audit in several phases. Each finding is checked independently and written in a format other tools can read. It was GitHub's top riser by stars on a day dominated by news of security flaws found with AI.

    GitHub Trending · Fastest risingRead the sourceProduct notes

  3. eMudhra's certificate authority keys were weak enough to crack in milliseconds

    A weak key at a trusted certificate authority puts every site and device that trusts it at risk.

    Security researcher Scott Helme explains how keys behind eMudhra's certificate authority, which issues HTTPS certificates, could be cracked in milliseconds. Mozilla has published how it is protecting Firefox users after the intermediate certificates were compromised. The story drew long threads on Hacker News and Lobsters.

    HN front page · Worth a debateRead the sourceDiscussion

  4. Recursive releases Flint, a coding model it says AI mostly built

    If labs can automate most of model development, new open models could arrive much faster.

    Startup Recursive released Flint, an open-weight coding model with 40 billion parameters. According to Wired, the company says its own AI system, the Eureka Machine, automated most of the development work. Recursive argues the benefits of using AI to build AI outweigh the risks.

    Techmeme · New ideaRead the sourceProduct notes

Go deeper: themes, product notes, every source

Themes across sources

6AI-found vulnerabilities · Agent supply-chain safety · Agent memory layers · Coding-agent control surfaces · Agent harnesses · Open-weight frontier models

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

AI-found vulnerabilities

New · 8 items from 4 sources

Two critical PHP flaws, nicknamed PHP2Shell, were found with OpenAI's GPT-5.6 Sol and patched in a rush that took European public websites offline. Anthropic says Claude flagged 493 security issues across 25 open-source projects, Cloudflare published a security-audit skill for coding agents, and a new group proposes coordinating fixes as agents find more bugs.

Why now. Frontier models can now read large codebases closely enough to find memory bugs that human review missed for years.

PHP2Shell: AI uncovers critical RCE in PHPHN front page

More on this

Agent supply-chain safety

Steady · 4 items from 2 sources

A compromised LiteLLM release on PyPI installed a credential stealer, a popular library many agent stacks depend on. Muse was found exposing secrets in shared conversations, and YC launches Alter and Armorer Labs sell access control and security checks for agents.

Why now. Agents hold API keys and act for several users, so one poisoned dependency or leaky share link exposes far more than before.

Compromised LiteLLM 1.107.0 on PyPI Installs a Credential Stealer and BackdoorHN front page

Agent memory layers

Rising · 6 items from 4 sources

GitHub's risers include graphify, which turns a codebase and its docs into a queryable knowledge graph, ai-memory for long-term memory across agent tools, and headroom, which compresses what reaches the model. Product Hunt adds a wiki that agents write, and a new paper keeps execution state as memory.

Why now. Long agent sessions keep hitting context limits, so remembering and compressing well now matters as much as model size.

Graphify-Labs/graphifyGitHub Trending

Coding-agent control surfaces

Steady · 4 items from 2 sources

Orca keeps climbing on GitHub as a place to run a fleet of coding agents, and Product Hunt brings ilha for usage limits in the Mac notch, BAND for agent teamwork and Agent Pets that live in Chrome. The common idea is a layer for watching and steering agents rather than another agent.

Why now. People now run several agents at once on subscriptions with usage limits, and need to see what each is doing.

stablyai/orcaGitHub Trending

More on this

Agent harnesses

Rising · 4 items from 3 sources

ponytail, a harness that pushes agents to write less code, was GitHub's top riser by stars. Mouse-1 sells a small decision model to plug into a harness, Harness squared targets self-improvement, and a paper asks which models should follow when a harness evolves.

Why now. The scaffolding around models now decides much of an agent's behaviour, and it can be tuned without retraining.

DietrichGebert/ponytailGitHub Trending

Open-weight frontier models

Rising · 4 items from 2 sources

DeepSeek V4.1 Flash topped Hugging Face's trending models, Meta's Muse Spark Open stays near the top, and Recursive released Flint, an open 40B coding model it says was developed mostly by its own AI system. Reflection's Beam is another large open release.

Why now. Open models close enough to the frontier let teams run strong coding and reasoning models on their own hardware.

deepseek-ai/DeepSeek-V4.1-Flash-1017HF trending models

Fading: Agent-ready Android (last seen 2026-10-20) · On-device AI (last seen 2026-10-19) · Verifying agent work (last seen 2026-10-19) · Agent sandboxes (last seen 2026-10-17) · Small decision models (last seen 2026-10-16) · AI-driven reverse engineering (last seen 2026-10-16)

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

cloudflare/security-audit-skill

repoGitHub Trending★ 1,997 today

What
A skill for coding agents from Cloudflare that runs a security audit in several phases and produces verified, machine-readable findings.
For whom
Teams that want their coding agent to review their own code for security problems.
Problem
Asking an agent to look for vulnerabilities gives long lists with many false alarms that are hard to act on.
Why now
The same week showed models finding serious real-world flaws, so teams want a repeatable way to do it.
Wedge
Independent verification of each finding and a structured output that tools can consume.
Money
Free; supports Cloudflare's security brand and products.
Like
Semgrep · CodeQL · Snyk

LessonPackaging an expert process as a reusable skill can travel further than a standalone product.

Recursive releases Flint, an open-weight, 40B-parameter coding model developed with its Eureka Machine AI system, which automated most of the development (Will Knight/Wired)

newsTechmemeHype flag

What
Flint, an open-weight 40B coding model from Recursive, which says its Eureka Machine AI system automated most of the development.
For whom
Teams that want a self-hostable coding model, and researchers watching AI-built AI.
Problem
Building a competitive model takes large research teams and long cycles.
Why now
Agents can now run many training experiments with little human input.
Wedge
The story of the process, AI building AI, as much as the model.
Money
Open weights as a showcase for the Eureka Machine system.
Like
Qwen Coder · Codestral · DeepSeek Coder

How much of the work was automated is the company's own claim and hard to verify.

LessonA product can be a proof for the tool that made it.

Graphify-Labs/graphify

repoGitHub Trending★ 1,434 today

What
A tool that turns a codebase with its docs, SQL schemas, configs and PDFs into a knowledge graph that coding agents can query, offered as a skill for several agent CLIs.
For whom
Developers working in large or unfamiliar codebases with a coding agent.
Problem
Agents rediscover how a codebase fits together in every session by reading files again.
Why now
Skills that work across Claude Code, Cursor, Codex and Gemini CLI make one tool reach many users.
Wedge
A single slash command that builds the graph.
Money
Open source, no model yet.
Like
Sourcegraph · XERJ · Aider repo map

LessonShipping as a skill for every major agent tool spreads faster than building yet another agent.

headroomlabs-ai/headroom

repoGitHub Trending★ 477 todayHype flag

What
A library that compresses tool outputs, logs, files and RAG chunks before they reach a language model.
For whom
Teams paying for large amounts of agent tokens, especially with JSON-heavy tools.
Problem
Agents send verbose tool output to the model, which costs money and fills the context.
Why now
Agent token spending is now large enough that a percentage saving is worth an extra library.
Wedge
The claim of the same answers with fewer tokens, which is easy to test.
Money
Open-source library; a hosted version is likely later.
Like
context-mode · LLMLingua

The 20% and 60 to 95% savings are the project's own figures and depend on the data.

LessonA measurable saving, here tokens, is the easiest pitch for infrastructure.

akitaonrails/ai-memory

repoGitHub Trending★ 303 today

What
Long-term memory for coding-agent CLIs that also helps hand work over between agents from different vendors.
For whom
Developers who switch between several coding agents on the same project.
Problem
Each agent keeps its own memory, so moving to another tool loses context.
Why now
Price and usage-limit changes push people to move between agent vendors more often.
Wedge
Vendor-neutral memory as the hand-off point between agents.
Money
Open source, no model yet.
Like
mem0 · Claude Code memory · graphify

LessonWhen users switch between rival tools, the neutral layer between them becomes valuable.

ilha

productProduct Hunt#5 on Product Hunt

What
A Mac app that shows your Claude Code and Codex usage limits live in the MacBook notch.
For whom
Developers on agent subscriptions who keep running into usage limits.
Problem
Limits are only visible inside each tool, so people hit them by surprise in the middle of work.
Why now
Subscription plans with rolling usage limits became the main way people pay for coding agents.
Wedge
One glanceable number in a spot that is always visible.
Money
Likely a small paid Mac app.
Like
iStat Menus · ccusage

LessonA new pricing model creates small but real needs, here tracking limits, that a tiny utility can fill.

Mouse-1

productProduct Hunt#6 on Product HuntHype flag

What
A decision model meant to plug into an agent harness.
For whom
Builders of agent harnesses who route steps between models or tools.
Problem
Using a large model for every small choice in a harness is slow and expensive.
Why now
Harnesses now make many small routing decisions per task.
Wedge
A specialised small model for one job inside the harness.
Money
Likely API usage.
Like
Martian router · RouteLLM

The listing does not say what decisions it makes or how it was evaluated.

LessonSmall models win by owning one narrow, frequent decision.

Triage by Chodeberg

productProduct Hunt#20 on Product Hunt

What
A tool that reproduces bugs described by users in clean sandbox environments.
For whom
Support and engineering teams that receive vague bug reports.
Problem
Engineers spend a lot of time trying to reproduce a bug from a short user description.
Why now
Agents can now follow written steps in a sandboxed app and record what happens.
Wedge
Starting from the user's own words rather than from logs.
Money
Likely per-seat or usage pricing for teams.
Like
Jam · Sentry session replay · Simulithic

LessonBug reproduction is a narrow, painful step that fits agents well because success is easy to check.

Simulithic — AI users that test your app on every device and browser.

productYC Launches25 votes

What
AI users that test your app on every device and browser.
For whom
Teams shipping code faster with agents and lacking test coverage across devices.
Problem
Verifying that code works lags behind how fast agents can now write it.
Why now
Agent-written code increased the volume of changes that need checking.
Wedge
Simulated users rather than written test scripts.
Money
Subscription or usage-based testing service.
Like
BrowserStack · Maestro · Firebase Test Lab

LessonFor Android teams: when writing code gets cheaper, the next bottleneck is checking it on real devices.

Launch HN: Alter (YC S25) – Access control for multi-user agents

discussionHN front page16 pts · 12 comments

What
Access control for agents that act on behalf of several users.
For whom
Companies deploying shared agents inside teams or for customers.
Problem
An agent shared by many people can use one person's access for another person's request.
Why now
Agents moved from personal assistants to shared team tools with real permissions.
Wedge
Focusing on the multi-user case that existing auth tools ignore.
Money
B2B security pricing.
Like
Auth0 · Oso · Permit.io

LessonEach new kind of actor, here a shared agent, needs its own answer to who is allowed to do what.

Veeza AI — AI tourist visa agent for people with weak passports

productYC Launches84 votes

What
An AI agent that handles tourist visa applications for people whose passports need visas for most countries.
For whom
Travellers from countries with weak passports facing slow, complicated visa paperwork.
Problem
Visa applications mean confusing forms, documents and appointments, often handled by paid agents.
Why now
Agents can now fill long forms and track requirements per country.
Wedge
Serving an underserved group that travel apps rarely design for.
Money
Likely a fee per application.
Like
VisaHQ · Atlys · iVisa

LessonPaperwork that people already pay human agents to do is a clear opening for AI agents.

deepseek-ai/DeepSeek-V4.1-Flash-1017

modelHF trending modelstrending 1,886

What
DeepSeek V4.1 Flash, a release of DeepSeek's open model that topped Hugging Face's trending models.
For whom
Developers wanting a fast, cheap open model to self-host or call through providers.
Problem
Strong closed models are expensive at scale and cannot be run on your own hardware.
Why now
Open labs now ship fast variants within weeks of their main models.
Wedge
A Flash version tuned for speed and cost.
Money
Open weights plus DeepSeek's own low-priced API.
Like
Qwen3.8 · Muse Spark Open · Gemini Flash

LessonReleasing a fast tier quickly keeps an open model in daily use, not only in benchmarks.

jqssun/android-airplay-server

repoGitHub Trending · KotlinAndroid★ 46 today

What
An open-source AirPlay receiver for Android that supports screen mirroring, video and audio.
For whom
People with Apple devices who want to cast to an Android TV, tablet or phone.
Problem
AirPlay normally needs Apple TV or licensed hardware, so mixed households cannot cast freely.
Why now
Android TV and tablets are common screens in homes that also use iPhones.
Wedge
Being the first open-source implementation, per the project.
Money
Open source, no model.
Like
AirScreen · Shairport Sync

LessonBridging two ecosystems that refuse to talk to each other is a durable kind of project.

shiaho777/web-to-app

repoGitHub Trending · KotlinAndroid★ 39 today

What
A toolkit that turns a website into an Android app, building the APK entirely on the phone.
For whom
Small site owners and hobbyists who want an app without a computer or Android Studio.
Problem
Wrapping a site as an app normally needs a development setup and signing know-how.
Why now
Phones can now build and sign APKs locally fast enough.
Wedge
Everything happens on the phone itself.
Money
Open source, no model.
Like
Bubblewrap · Capacitor · Median

LessonFor Android developers: removing the computer from a workflow opens it to many more people.

rikkahub/rikkahub

repoGitHub Trending · KotlinAndroid★ 41 today

What
RikkaHub, an Android app that connects to many language model providers from one interface.
For whom
Android users who use several AI providers or bring their own API keys.
Problem
Each AI provider has its own app, so switching models means switching apps.
Why now
Model choice changes monthly, and many providers offer compatible APIs.
Wedge
One native Android client for many providers.
Money
Open source, no model.
Like
ChatBox · Open WebUI · LibreChat

LessonA neutral client gains users whenever the services behind it compete hard.

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

12
  • cloudflare/security-audit-skillnewcard

    A coding-agent skill for multi-phase security audits with independently verified, machine-readable findings

    ★ 1,997 today

  • vllm-project/vllmnew

    vLLM, the standard open-source engine for serving language models, keeps gaining stars as self-hosting grows.

    ★ 439 today

  • stablyai/orca

    Orca lets you run a fleet of coding agents on your own subscriptions from desktop or phone.

    ★ 1,543 today

  • headroomlabs-ai/headroomcard

    Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.

    ★ 477 today

  • DietrichGebert/ponytailnew

    A harness that tells agents to avoid writing code they do not need, the top GitHub riser by stars.

    ★ 2,051 today

  • Comfy-Org/ComfyUI

    ComfyUI, the node-based interface for image and video models, stays a default tool for local generation.

    ★ 914 today

  • supabase/supabasenew

    Supabase keeps trending as the Postgres backend many AI-built apps start on.

    ★ 509 today

  • twentyhq/twentynew

    An open-source CRM pitched as a Salesforce alternative designed for AI.

    ★ 105 today

Show 4 more
  • shiyu-coder/Kronos

    A foundation model trained on the language of financial market data.

    ★ 570 today

  • strelov1/freehirenew

    An open-source job search engine built for job seekers rather than recruiters.

    ★ 396 today

  • akitaonrails/ai-memorycard

    Solution for long term memory for agent coding CLIs and to facilitate handoff between different agent vendors

    ★ 303 today

  • Graphify-Labs/graphifynewcard

    Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, every edge explained, no vector store.

    ★ 1,434 today

Product Hunt

50
  • Burrow MacOSnew

    Runs Windows and Linux on a Mac for free.

    #1 on Product Hunt

  • UseIt Studionew

    A design agent that edits real, editable design files rather than images.

    #2 on Product Hunt

  • LearnExcelnew

    Learn Excel by typing real formulas, graded instantly

    #3 on Product Hunt

  • Flyweelnew

    Ad spend cards and funding with an agent audit over MCP, finance tools exposed to agents.

    #4 on Product Hunt

  • ilhanewcard

    Your Claude Code and Codex limits, live in the notch

    #5 on Product Hunt

  • Mouse-1newcard

    A decision model for your agent harness.

    #6 on Product Hunt

  • BAND for Developersnew

    Team collaboration features for coding agents.

    #7 on Product Hunt

  • The Brainsnew

    Founders and investors packaged as MCP servers with cited answers.

    #8 on Product Hunt

Show 42 more
  • AskDoozynew

    An AI coworker in Slack that does your real work

    #9 on Product Hunt

  • Tendnew

    Personal AI for businesses and entrepreneurs.

    #10 on Product Hunt

  • Intelli for ChatGPT & Claudenew

    Run WhatsApp, Instagram & web chat from ChatGPT or Claude

    #11 on Product Hunt

  • Agent Petsnew

    Shows coding agents as pets on every page in Chrome, a playful status display.

    #12 on Product Hunt

  • The Librarynew

    Your business context organized by queries

    #13 on Product Hunt

  • Koinew

    Every tool you need, in one Mac app

    #14 on Product Hunt

  • PrivOSnew

    Chat, Docs, Kanban Tasks and Al agents that act, self-hosted

    #15 on Product Hunt

  • Nimbalyst Wikinew

    A wiki that your agents write for you.

    #16 on Product Hunt

  • Notion Autonew

    Notion's rebuilt agents run for hours at lower cost, a sign of long-running agents in mainstream apps.

    #17 on Product Hunt

  • Blooionew

    Apple Messages for Business made easy

    #18 on Product Hunt

  • Yaritenew

    A daily list of who to call that keeps all data on your phone.

    #19 on Product Hunt

  • Triage by Chodebergnewcard

    Reproduce user described bugs in clean sandbox environments

    #20 on Product Hunt

  • NativeShipnew

    The easy way to launch your n8n native integration

    #21 on Product Hunt

  • Flagsmith Experimentsnew

    Open-source A/B testing built on existing feature flags.

    #22 on Product Hunt

  • Needlenew

    The agent-native outbound engine

    #23 on Product Hunt

  • Woven Videonew

    Edit any video by chat, on a real Mac timeline

    #24 on Product Hunt

  • TwinMind for Desktopnew

    Transcribe meetings and dictate anywhere on Mac & Windows

    #25 on Product Hunt

  • iglassnew

    Brings the Liquid Glass look to iTerm2 and Ghostty terminals.

    #26 on Product Hunt

  • Scout by Fininew

    The AI coworker for support teams, on Slack

    #27 on Product Hunt

  • Lea by LeanAIOSnew

    Autonomous AI team for construction management & pre-con

    #28 on Product Hunt

  • Animatica Marionettenew

    Pose humanoid rigs in Blender by dragging hands and feet

    #29 on Product Hunt

  • hibinew

    An open-source alternative to Obsidian for notes.

    #30 on Product Hunt

  • Hotlynew

    Get paid to answer your DMs. Open source.

    #31 on Product Hunt

  • Sorcerernew

    Simulate an AI agent community reacting to "what if"

    #32 on Product Hunt

  • ESP-Mosaiconew

    A modular ESP development board designed to be built with by agents.

    #33 on Product Hunt

  • OpenShip Cloudnew

    Managed Cloud infrastructure that you never locked in

    #34 on Product Hunt

  • Praexnew

    A coding agent that runs on its own models rather than third-party APIs.

    #35 on Product Hunt

  • Glance for Macnew

    Your MacBook notch, now a live, cinematic island

    #36 on Product Hunt

  • Tiny Little Officenew

    A tiny office suite where the link is the document

    #37 on Product Hunt

  • Portlivonew

    Lists every local dev server on your Mac in the menu bar.

    #38 on Product Hunt

  • Intel® GPU D-seriesnew

    Intel GPUs aimed at memory-heavy AI inference without expensive HBM memory.

    #39 on Product Hunt

  • Lugh Xtudionew

    The video studio where AI agents edit with you

    #40 on Product Hunt

  • RunInfranew

    Open models, built for agents.

    #41 on Product Hunt

  • Jotform App Builder for ChatGPT & Claudenew

    Build and edit no-code apps inside ChatGPT & Claude

    #42 on Product Hunt

  • Stellanew

    Your emotional wellness AI. A friend to figure it out with.

    #43 on Product Hunt

  • Kling 4.0new

    Kling 4.0, a faster and cheaper AI video model for frequent creators.

    #44 on Product Hunt

  • call4menew

    Gives an AI agent a phone number that can make and answer calls.

    #45 on Product Hunt

  • /dev/nullnew

    a social network for devs

    #46 on Product Hunt

  • bondsnew

    Every friends' chat gets an AI friend with superpowers

    #47 on Product Hunt

  • Harness²new

    A harness built around making agents improve themselves.

    #48 on Product Hunt

  • ToDo.isnew

    Write a to-do and your AI agent gets it done

    #49 on Product Hunt

  • Stagenew

    Review, fix and sign off working prototypes and products

    #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

19
Show 11 more

HF trending spaces

15
Show 7 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

366 items from 77 sources, 5 of them seen in several sources. Every source was fetched.

Source health

77Which sources were fetched today, and how many items each returned.
SourceSectionItemsStatus
GitHub TrendingMain12Fetched
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 & releases15Fetched
Android Developers BlogAndroid · Official & releases4Fetched
Android Studio releasesAndroid · Official & releases3Fetched
Kotlin blogAndroid · Official & releases3Fetched
Kotlin releasesAndroid · Official & releases0Fetched
Compose Multiplatform releasesAndroid · Official & releases4Fetched
Android Developers YouTubeAndroid · Official & releases6Fetched
Kotlin on YouTube (JetBrains)Android · Official & releases2Fetched
Google Developers BlogAndroid · Official & releases1Fetched
AndroidX merged API changesAndroid · AOSP / AndroidX insider24Fetched
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 · Experts1Fetched
CommonsWare (Mark Murphy)Android · Experts2Fetched
TechYourChance (Vasiliy Zukanov)Android · Experts1Fetched
Kt. Academy (Marcin Moskała)Android · Experts2Fetched
Touchlab (Kotlin Multiplatform)Android · Experts2Fetched
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 releases0Fetched
Molecule releasesAndroid · Library releases0Fetched
Turbine releasesAndroid · Library releases0Fetched
Paparazzi releasesAndroid · Library releases0Fetched
SQLDelight releasesAndroid · Library releases0Fetched
OkHttp releasesAndroid · Library releases0Fetched
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 releases0Fetched
detekt releasesAndroid · Library releases0Fetched
ktlint releasesAndroid · Library releases0Fetched
MockK releasesAndroid · Library releases0Fetched
Lottie releasesAndroid · Library releases0Fetched
Authority Insights (leaks & teardowns)Android · Leaks & ecosystem7Fetched
9to5Google · AndroidAndroid · Leaks & ecosystem8Fetched
Android AuthorityAndroid · Leaks & ecosystem6Fetched
Android FaithfulAndroid · Leaks & ecosystem2Fetched
Android WeeklyAndroid · Community15Fetched
GitHub Trending · KotlinAndroid · Community8Fetched
Lobsters · androidAndroid · Community2Fetched
HN · AndroidAndroid · Community7Fetched
ProAndroidDevAndroid · Community5Fetched
Kotlin WeeklyAndroid · Community1Fetched
Philipp Lackner (YouTube)Android · Community2Fetched
1 of 5

AppFunctions lets your app offer actions to on-device AI

This is how your app's features could become something an assistant can do for users without opening your app.

A new explainer walks through AppFunctions, an Android platform API with a Jetpack library. You mark suspend functions with an annotation, and agents such as Gemini can discover and call them on the device, much like tools on an MCP server. It needs Android 16, it is still an experimental preview, and Gemini access is limited to testers for now.

ProAndroidDev · CommunityRead the source

1 of 4

AI finds two critical PHP flaws that affect much of the web

AI-found bugs in core infrastructure are arriving faster, and patching windows are getting shorter.

PHP's maintainers patched two critical flaws, nicknamed PHP2Shell, that let attackers run code on servers without logging in. OpenAI says its GPT-5.6 Sol model found them in August. Fixes cover only PHP 8.2 and newer, Cloudflare added firewall protection before disclosure, and some European public websites went offline while teams patched.

HN front page · Across sourcesRead the sourceDiscussion