# Shuriken 1.0: why we called it 1.0 after five days > Shuriken 1.0 shipped on 4 October 2026, five days after the first prerelease. Why we called it 1.0, what changed, and the 1.1 that followed that evening. Source: https://shuriken.sh/blog/shuriken-1-0 Shuriken 1.0 shipped on 4 October 2026, five days after the first prerelease. Why we called it 1.0, what changed, and the 1.1 that followed that evening. Update By Shuriken Updated 27 October 2026 3 min read **Contents** Shuriken 1.0.0 was tagged on 4 October 2026 at 08:03, Amsterdam time. That is five days after the first prerelease on 29 September, and one night after three releases in one day. Five days isn't long, and we knew it. ## Why 1.0, and why so early # To us a 1.0 means one thing: you can rely on it. By the morning of 4 October the core loop held up. You add the accounts you pay for, start chats with your agents on your own computers, keep going on the next account when one reaches its limit, and finish the work into a branch or a pull request. We were already running that loop every day on our own projects. So 1.0 meant that loop works. It didn't mean finished. It was early, with rough edges, a long list of features still to come and several new versions a week after it. But the thing people came for worked, and calling it 0.13 would have been modesty for its own sake. ## What changed in 1.0 # ### Chats run only on accounts you added # In 1.0 a chat runs only on accounts you added to Sources, never on some sign-in that happened to be lying around on the device. It sounds like a small rule, but everything else depends on it. You know which accounts Shuriken can use because you put them there, and [automatic source switching](https://shuriken.sh/features/source-switching) moves between exactly those. ### The queue is one strip on the composer # Messages you type while the agent works wait in the queue. In 1.0 the queue became one strip on the composer, one line per task, right where you're typing. You see what's waiting without leaving the chat. The queue as it looks today: one strip on the composer, one line per task. ### Finish says where the work goes # Finish runs your checks, then merges or opens a pull request. In 1.0 it says where the work is going before it goes. A repository with no remote merges into main, and the button tells you so. ### Spend counted by the account each turn ran on # When a chat moves between accounts in the middle of a task, its cost now follows the work: spend is counted by the account each turn ran on, once. A stopped turn, a failed turn and a mid-turn account move all land on the right account. ### The Mac app installs in one line # One line in Terminal installs the Mac app. The download page puts that line first, with the way past Gatekeeper for the .dmg right under it. ### The agent limit is a slider # The per-device Agents at once setting from 0.11 became a slider up to Unlimited, with a number field if you'd rather type. Unlimited is still the default. Shuriken runs your coding agents on your own computers and keeps them going when an account hits its limit. [Try Shuriken](https://shuriken.sh/) ## 1.1.0, the same evening # At 21:31 the same day, 1.1.0 fixed what turned up within hours. If you stopped a turn and then sent the next message, the stop now clears and the message goes. When none of your Sources can run a chat's agent, Send is greyed out and says why, instead of failing after you press it. While the agent is writing, the button turns into a plus that adds your message to the queue, so there's always a way to say the next thing. And Shuriken's helper processes now stop when you quit it, instead of humming on after you thought you'd closed it. (Since 3.2 you can choose to keep agents running with the window closed, through the background service. That one's on purpose, and your call.) ## What came next # As for what 1.0 didn't have yet: the next day's release, 1.2, brought [Import from Claude Code and Codex](https://shuriken.sh/blog/claude-and-codex-together), so your projects came with you instead of starting from zero. Their past chats followed in 1.5. If you'd like the version that has had several weeks more work on it, [try Shuriken](https://shuriken.sh/). Share [X](https://x.com/intent/post?url=https%3A%2F%2Fshuriken.sh%2Fblog%2Fshuriken-1-0&text=Shuriken%201.0%3A%20why%20we%20called%20it%201.0%20after%20five%20days)[LinkedIn](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Fshuriken.sh%2Fblog%2Fshuriken-1-0)[Hacker News](https://news.ycombinator.com/submitlink?u=https%3A%2F%2Fshuriken.sh%2Fblog%2Fshuriken-1-0&t=Shuriken%201.0%3A%20why%20we%20called%20it%201.0%20after%20five%20days)[Reddit](https://www.reddit.com/submit?url=https%3A%2F%2Fshuriken.sh%2Fblog%2Fshuriken-1-0&title=Shuriken%201.0%3A%20why%20we%20called%20it%201.0%20after%20five%20days)[E-mail](mailto:?subject=Shuriken%201.0%3A%20why%20we%20called%20it%201.0%20after%20five%20days&body=https%3A%2F%2Fshuriken.sh%2Fblog%2Fshuriken-1-0) ## Frequently asked questions **When was Shuriken 1.0 released?** On 4 October 2026 at 08:03 Amsterdam time, five days after the first prerelease on 29 September. Shuriken 1.1.0 followed the same evening. **Does 1.0 mean Shuriken is finished?** No. It meant the core loop was solid enough to rely on: add your accounts, run agents, switch when a limit hits, finish the work. A lot has shipped since. **Why is Send greyed out in a chat?** Since 1.1.0, Send is greyed when no account in your Sources can run that chat's agent, and it tells you why. Add an account for that provider to continue. Shuriken The harness for your coding agents. Run Claude Code, Codex and Gemini CLI on your own computers, move chats to your next account at a limit, and follow every chat from any device. [Try Shuriken](https://shuriken.sh/) [How it works](https://shuriken.sh/features/source-switching) ## Keep reading - [How Shuriken started: a tool we built for ourselves: Story 27 Oct 2026 4 min read We build software for clients and ourselves. No harness fit how heavily we use AI, so we built our own. Then a programmer asked: why don't we sell this?](https://shuriken.sh/blog/how-shuriken-started) - [What's new in Shuriken 5.1: Update 27 Oct 2026 16 min read Shuriken 5.1 adds sign-in with Google, Apple or an email code, passkeys, Allow and Always allow at a MacBook's notch, and billing and fuel in Sources.](https://shuriken.sh/blog/shuriken-5-1) - [Test-driven development with Claude Code: Story 27 Oct 2026 4 min read Claude stops when work looks done, so give it a failing test to run. Our TDD prompt in three steps, and a Stop hook that runs tests when Claude finishes.](https://shuriken.sh/blog/tdd-with-claude-code) - [Why chats said "Connection lost", and the fix: Story 27 Oct 2026 4 min read On 5.0.0, one very long chat could freeze Shuriken's background service: "not reached" blinking, window reloads, stopped agents. The cause and the fix.](https://shuriken.sh/blog/why-chats-said-connection-lost) - [How to change the model in Claude Code: Guide 27 Oct 2026 3 min read Change the model in Claude Code with /model, claude --model, an environment variable or settings.json. /model wins; Enter in its picker sets your default.](https://shuriken.sh/guides/change-model-in-claude-code) ## Free tools - [Claude limit reset tracker](https://shuriken.sh/tools/limit-resets/claude) - [Codex limit reset tracker](https://shuriken.sh/tools/limit-resets/codex)