Hi team!
If you build for Fire TV, check out our two big testing announcements: Appstore Beta Hub lets your Live App Testing testers install and manage beta apps straight from their Fire TV, and accelerated subscriptions take a subscription renewal test from a full plan period down to minutes. On the Vega side, we have a Developer Bulletin on Vega DevKit after the Mac OS 27 upgrade.
If you've been active online at all the past couple weeks, you must have heard about Jev! If you've been curious but don't know where to get started, check out the explainer videos from Neha and Mike in the Dev Posts section. We also have a bunch more AI content for you, including stuff on Strands, Bedrock, and Bee.
One more post I wanted to highlight, Claude now has a developer portal of their own, where you can submit and download plugins! If you're interested, check out their blog on how to use it.
A ton of the Amazon Developer team will be in Berlin next week for Next App Dev Con Berlin, including a workshop from Anisha and Gio, a panel with Anisha, and a lightning talk from me! Let me know if you'll be there and come say hi!
⚛️ Developer Bulletin: Issues with Vega DevKit after Mac OS 27 upgrade — upgrading a Mac to OS 27 removes Rosetta, which several DevKit components depend on. One command restores it:
softwareupdate --install-rosetta --agree-to-license.📢 Claude Opus 5.5 is now available on AWS — generally available through both Amazon Bedrock and Claude Platform on AWS, built to keep reporting on its progress and to finish multi-step coding work in fewer tokens than Opus 5.
🧪 Appstore Beta Hub
- Your Live App Testing testers no longer need a sideload or an emailed link: Appstore Beta Hub is a self-service app they install from the Amazon Appstore on a supported Fire TV, and they manage their beta apps from the device itself.
⏳ Accelerated Subscriptions for Live App Testing
- Testing a subscription renewal used to mean waiting out the full plan, so an annual plan meant an annual wait. Accelerated subscriptions compress that to minutes. Worth knowing before you switch: enabling it cancels all your existing subscription testing.
🔥 Best Buy and Amazon expand Fire TV partnership, opening ad inventory to Best Buy Ads
- If you ship a Fire TV app, the installed base you're targeting just got quantified: Best Buy's renewed multi-year deal keeps Fire TV as the only software on every Insignia TV it builds, and the announcement puts numbers on that surface at tens of millions of Fire TV sets across 100-plus models from 24 to 85 inches, with the ad-inventory piece advertiser-facing and rolling out in February 2027.
🎬 Prime Video adds on-demand news clips to its news destination
- Prime Video's free news destination now carries News Clips, curated playlists that stitch short-form clips from multiple networks into Trending Topics and Local News carousels, alongside the 24/7 live feeds already there. Living-room devices get it first with mobile and web to follow, which is the device-tier rollout order to expect when you ship a new surface into the Prime Video app.
Our Build, Ship, Shape: Amazon Developer Hackathon is underway online, so start building before the deadline. In person, Next App Dev Con Berlin is our big one this season, with Developer Advocate Anisha Malde on a panel and co-leading a workshop with Giovanni Laquidara, plus a lightning talk from me. I'm also giving a talk at CityJS Athens. We hope to see you there!
| Date | Event | Location | Description |
|---|---|---|---|
| Aug 31 – Oct 23 | Build, Ship, Shape: Amazon Developer Hackathon | Online | Our first global developer hackathon: build on Fire TV, Alexa+, Ring, or Bee across four tracks for a share of a $190K pool of cash and AWS credits, with submissions closing October 23. |
| Oct 7–9 | Next App Dev Con Berlin | Berlin, DE | Dev Advocate Anisha Malde joins a panel and co-leads a workshop with Giovanni Laquidara, and Moses Roth gives a lightning talk. Mobile and app-platform conference at Berlin's CityCube, with workshops, talks, and platform announcements. |
| Oct 21–23 | CityJS Athens | Athens, GR | Dev Advocate Moses Roth is giving a talk. Athens edition of the CityJS conference series, covering modern JavaScript and frontend frameworks. |
| Oct 21–22 | NAB Show New York | New York City, US | East-coast edition of NAB Show — broadcast, media, and entertainment-tech industry conference, at the Javits Center. |
| Oct 22–23 | AGNTCon + MCPCon North America | San Jose, US | Linux Foundation conference for the open agentic AI ecosystem, with MCP and agent-systems tracks for builders, contributors, and enterprises. |
| Oct 23–26 | React Advanced London | London, GB | Annual advanced-React conference in London for senior practitioners working with the framework at scale. |
| Oct 28–29 | GitHub Universe | San Francisco, US | GitHub's flagship two-day event for developers, maintainers, and technical leaders, focused on collaboration and workflows in the agentic era. |
| Oct 29–31 | React India | Goa, IN | India's largest React conference, covering web, mobile, and emerging React platforms. |
| Nov 9–12 | Web Summit | Lisbon, PT | One of Europe's largest tech conferences, with broad coverage across AI, devices, startups, and platforms. |
| Nov 10–12 | AI Engineer Code Summit | San Francisco, US | The code-focused summit in the AI Engineer conference series, on agent and LLM development. |
Future of Software Development | Closing Panel at Agent Conf 2026 is the one to start with, a closing-panel conversation about where the craft goes next. Then What does a million tokens actually buy? puts numbers on a question most of us have been eyeballing, and Stop Starting From Scratch: Build Agents with Strands harness is the practical follow-on if you want to build rather than read.
📽️ Future of Software Development | Closing Panel at Agent Conf 2026
Chris Trag moderates Agent Conf's closing panel with Margaret Pieczkowski, Grzegorz Wołodko, Bartek Pucek, and Carl Ross on the bottleneck that shows up once agents write most of your code and other realities of the new software development world we now live in.
📰 Bee Private Compute: How Bee keeps your conversations private
A confidential-computing architecture worth reading as a reference design: Bee transcribes audio on device in real time and discards it, then encrypts transcripts, summaries, and to-dos under keys the device alone creates and holds, so Amazon has no master key. Before anything leaves the device it attests that the receiving server is running approved software, and processing happens inside an isolated enclave built on the AWS Nitro System and NitroTPM, the same infrastructure that backs Nitro Enclaves for AWS customers.
📽️ What Can You Build with Jev? Smart AI Agent Routing
Neha Sharma demos intent-based routing in Jev from TypeSafe AI, sending a single prompt to four specialist agents (Researcher, Skeptic, Comedian, Optimizer) and showing the probability each one scores before the router picks a winner.
📽️ Meet Strands harness: Build State-of-the-Art Agents at 28% Lower Token Cost
Strands harness is AWS's new agent harness: it ships the plumbing you'd otherwise hand-tune for every agent, from context management and memory to tools and task tracking, behind a single call. The video builds a docs-maintainer bot that keeps documentation current on its own, and Morgan Willis walks the same build in the written version. Worth your time if you've been assembling that scaffolding yourself for every new agent.
📽️ Jev: Is This Tool Call Ready?
Mike Chambers brings together the two previous videos: Jev and Strands. Here Jev sits inside a Strands agent and checks a tool call before it runs, so a bad one is caught instead of executed. Watch this one for the pattern you'd actually drop into your own agent loop.
📰 How to Use Builder Center to Land Your Next Role
By Rufeda Ali
On the job hunt?
Rufeda gets into the practical case for treating your Builder Center profile like a résumé you send when someone asks what you've built, and the type of posts that actually earn a reply.
📰 What does a million tokens actually buy?
A price quoted per million tokens isn't comparable across providers, because a token isn't the same unit from one model to the next. Edvin actually measures the gap rather than guessing at it, and the same text can cost meaningfully more on one model than another purely because of how it tokenizes. Read it before you pick a model on the sticker price, especially for text-in-text-out work like summarization, extraction or RAG.
📰 Use open weight models as your AI coding agent with Amazon Bedrock
By Aris Tsakpinis, Sainath Miriyala, Saurabh Trikande, and Suryansh Rana
If your prompts and your code can't leave your own AWS account, this walkthrough points OpenCode at open weight models hosted on Amazon Bedrock so the whole coding agent runs inside your account. Multi-model routing is the part worth studying: heavy reasoning and quick edits bill per token against different models instead of one flat rate. Read it if data residency or a per-token budget is what has kept a coding agent off your stack.
📽️ Getting Started with Kiro Web and Cloud Sessions
Catherine Cao follows a single task end to end in Kiro Web, kicking off the work from the browser, handing it to an isolated cloud environment where Kiro runs it, and landing the result as a pull request. If you've only used Kiro on the desktop, the cloud session model is the part worth studying: your machine isn't the execution environment, so the task keeps running independently of your editor and you review the diff the same way you'd review any teammate's PR.
📰 Build plugins for Claude with the directory submission portal
Anthropic opened a submission portal for the Claude plugin directory, so an MCP connector or Agent Skill you've built can be packaged as a plugin, submitted for review, and tracked through the pipeline until it goes live. If you've wrapped an Amazon developer surface in an MCP server, this is the distribution path that puts it in front of Claude users without asking them to configure a local server by hand.
📬 Feedback
Questions or comments? Have an item for next week's post? Let me know!
Top comments (0)