A big one: a turn can now keep going when its model gives out, you can roll back what a turn did to your files, and bots can build up skills of their own.

• Fallback models. When a model hits its usage limit, loses its login, is unavailable or runs out of context, the turn can continue on the next model in your list, API-key providers included. Set it up in Settings Models and agents Fallback models. It's off until you add a model, and you can pin a conversation to its own model.

• Undo a turn's file changes. When a turn changes files, its answer shows the diffs, and "Undo changes of this turn" puts them back, whether the agent used its file tools or ran commands. You review each file first, can undo just one, and can redo. It works with or without Git.

• Bots keep their own skills. A bot can save a procedure it learned as a skill and improve it over time. In the bot's Settings Skills you see each skill's history, restore a version, edit or lock it, or require your approval before a change takes effect.

• One account rotation per provider. Claude and Codex accounts now share a single ordered rotation in Settings Accounts. An account you pick directly stays fixed and stops when it runs out instead of switching. The failover routes you already had become the new rotation.

• Long conversations hold up better. Agents search past messages by what they're about, not only by exact words, and compaction keeps decisions, pending requests and exact identifiers in fixed sections, with the latest part of the conversation kept word for word.

• Sturdier with local models. Ollama, LM Studio, llama.cpp and vLLM report the context window they actually run with, and malformed tool calls from any model are repaired before they fail.

Release notes: https://github.com/antonioducs/m...

This release is about running Maestrly on several Claude or Codex accounts without switching between them by hand.

• Active account in any model selector. Pick Claude Active account or Codex Active account and Maestrly runs on whichever of your accounts is active. You change it in one place: Settings Accounts Active account.

• Use each account until it runs out. Maestrly stays on an account until its quota ends, then moves to the next one and keeps using it, so the reset windows start one after another instead of overlapping. Or choose Return to the active account to go back to it as soon as it resets.

• Everything follows it. Chats, subagents, compaction and image interpretation all use the active account. Picking a specific account still works as before.

• Claude plan renames no longer break your sessions. Conversations and usage limits keep working when Anthropic renames a plan, including Max to Ultra. Existing Claude sessions are rebuilt once from the conversation history the next time you use them, and the history is kept.

• Long conversations switched to Google AI keep your whole message. Maestrly summarizes an oversized history before sending it instead of silently cutting off the end of the prompt. The first response after the switch can take a bit longer.

• Interrupted responses stop cleanly. If Maestrly closes or restarts mid-response, tools and subagent cards no longer spin forever, and the response says it was interrupted.

Release notes: https://github.com/antonioducs/m...

Bots can now hand their heavy work to your other computers, and wait for each other's answers without dropping what they were doing.

• Runners. Turn one of your computers into a runner and your bots can send it heavy jobs like video renders, model inference, large builds or test suites, instead of using the CPU and memory they share with their browsers. The outputs land back in the bot's folder and the conversation picks up with the result.

• Watch and cancel jobs. You see each job's CPU, memory and live log, and can cancel it from any paired computer. Jobs run natively on macOS and Linux, or in a Docker container on macOS, Linux and Windows.

• Easy to set up. On the computer, turn on Run jobs for bots in Settings Runner, then allow the bot in its Settings Runners. A linked bot can run any command on that computer, so only link bots you trust.

• Bots that wait for each other. When a bot asks another bot something, it can now wait for the answer in the same turn and finish the task, instead of reading the reply only on its next turn.

• Steadier long sessions. An approved plan survives context compaction, and Claude no longer claims it can't see an image you just attached after its session is rebuilt.

Runners need the computer, the bot server and the bot environment on v0.30.0, and the mobile app doesn't show them yet. If you run your own bot server, back up the gateway volume before updating it.

Release notes: https://github.com/antonioducs/m...

This round turns artifact pages into simple forms and makes bot logins stick around.

• Collect responses on an artifact. A page can now save one response from each person who opens it, like the answers to a form you send someone. Guests type their name once before their first response.

• Ask an agent for a form. Agents know the new page API and read the responses with the artifact_responses tool, so asking one to build a form just works.

• Manage what comes in. The artifact's details list the responses, delete them and export them as CSV or JSON. Share Responses turns them off.

• Bot browser stays open. A bot's Chromium no longer closes after 15 idle minutes, so it no longer signs the bot out of Google. It stays open until the bot pauses or its environment restarts.

• Logins survive reopens. Sites that keep their login in session cookies stay signed in when a bot's Chromium reopens after a pause or a restart.

Responses need a bot server on v0.29.0 or later. After updating, sign your bots in to Google once more.

Release notes: https://github.com/antonioducs/m...

This release makes remote and multi-device setups feel more like one place, and keeps your bots signed in to Google.

• Groups for remote workspaces. Organize the workspaces of a remote host or a paired computer into groups, just like on your own machine, with drag and drop, Move to group and New group. The groups live on the host, so everyone who uses it sees the same ones, and changes show up right away on other computers.

• Seen state synced across devices. Open a routine run or read a bot-to-bot conversation on one computer or phone, and it no longer shows as new on the others. Notifications and the bot's red dot follow along.

• Bots stay signed in to Google. A bot no longer has to sign in again after its environment restarts, updates or stops.

• Google sign-in in the drawer browser works. It no longer ends on "This browser or app may not be secure".

Synced seen state needs a bot server on 0.28. If you run your own, back up the gateway volume before updating.

Release notes: https://github.com/antonioducs/m...

Maestrly v0.27.0 is out with system notifications, interactive bot screen control on mobile, individual bot restarts, and more reliable browser sessions.

• System notifications and unread badges: Desktop and server conversations notify you when they finish a turn, fail, ask a question, or bring a plan, and bots notify you when they finish what you asked, complete a tracked routine run, or need your help. Clicking a notification takes you straight to that conversation or run, and unread dots on the Bots tab and app icon show what is waiting for you.

• Control a bot's screen from your phone: When a bot asks for help with a password, 2FA code, or CAPTCHA, you can now open its live screen in the mobile app, tap, scroll, and type with your phone's keyboard, and hand the screen back with a note so the bot carries on.

• Restart individual bots: You can now restart a single bot that shares an environment without restarting the whole environment or interrupting the other bots, keeping its conversation, queue, files, memory, and settings intact.

• Persistent browser sessions and Google sign-in: A bot's Chromium now saves its latest cookies cleanly when closing after idle time, pausing, or restarting so it stays signed in to sites like Google and YouTube, and bots can sign in to Google accounts again without automation blocks.

• Background VS Code updates and remote folder picker: The Code tab now downloads new VS Code versions in the background instead of making you wait when opening it, and you can browse and choose the projects folder on remote computers and server environments directly from Settings or chat cards.

Release notes: https://github.com/antonioducs/m...

Maestrly v0.26.0 gives every bot its own real Chromium browser on its desktop and makes browser automation much more dependable.

• Dedicated Chromium per bot: Each bot now runs a real Chromium window on its desktop with its own isolated profile, cookies, and site logins instead of an embedded browser. Sign-in pages and bot checks see an ordinary Chromium, you can use it directly from the bot's screen, and each conversation manages its own tabs (closing after 15 idle minutes and reopening on next use).

• Smarter browser snapshots and actions: Browser tools re-verify snapshot elements before clicking or typing, automatically scroll off-screen elements into view, and now read inside open shadow roots and same-origin frames.

• Clean recovery from unresponsive pages: Browser actions and scripts no longer hang indefinitely when a page stops responding, timing out cleanly and restarting the stuck page on the next navigation or reload.

• To-do continuity after context compaction: Agents keep working from their active to-do list after a conversation's context is compacted instead of leaving the last card frozen in the chat.

Release notes: https://github.com/antonioducs/m...

Maestrly v0.25.0 is out with direct peer-to-peer computer pairing, support for multiple bot servers, and image generation powered by Google AI.

• Direct computer-to-computer pairing: Connect your computers directly over your local network or Tailscale with a 6-digit code in Settings Computers, with no server in between. Switch machines from the machine switcher to run chats and workspaces on another computer using its own folders, terminals, accounts, and keys.

• Multiple bot servers: Pair your app with more than one bot server at the same time. Switch between servers in the Bots tab, access each server's workspace environments from the machine switcher, view artifacts from all paired servers in one place, and choose where new artifacts are published.

• Google AI image generation: Draw images with the image generation tool using a Google AI subscription and Gemini's image model in conversations on any model, or pick your preferred generator (Automatic, ChatGPT, or Google AI) in settings.

• Safer bot moves and easier recovery: Moving large bots now checks the 32 GB package limit before freezing so oversized bots keep running, and clicking Start on a running environment now restarts any stopped or failed bots inside it, including from the mobile app.

Release notes: https://github.com/antonioducs/m...

Maestrly v0.24.0 brings file management for bots across desktop and mobile, full mobile access to routine runs and bot-to-bot chats, and a smarter Git initializer for existing folders.

• Bot file manager on desktop and mobile: Open the new Files drawer from any bot conversation or the mobile app to browse, search, filter, bulk-download, or delete shared files across a single chat or the whole bot, configure its storage quota (1 GB to 20 GB), and choose whether deleting a routine run or bot-to-bot chat also removes its shared files.

• Routine runs and bot-to-bot threads on mobile: View a bot's Runs and Bots tabs right from the mobile app, read everything each run or thread did, reply directly, let paused bot-to-bot chats continue, or cite a run in your next message.

• Bot-started workspace chats on mobile: See and control workspace conversations that a bot started from your phone, with options to pause or resume the bot, release the chat to type in it yourself, and hand it back when you are done.

• Initialize Git with your existing files: Initializing Git in a folder without history now commits your existing project files instead of creating an empty commit, so worktree conversations start with your code ready while automatically excluding ignored files, dependencies, caches, large files, and likely secrets.

• Cleaner bot-to-bot exchanges and reliable Stop: Bots talking to each other now stop once the task is settled instead of replying to thank-yous in a loop, and stopping a bot turn while it is still preparing or compacting holds queued messages until you write again.

Release notes: https://github.com/antonioducs/m...

Maestrly v0.23.0 is out with parallel task execution and smart model delegation for bots, full plan reviews on mobile, and better subscription controls.

• Parallel bot tasks: Bots can split independent work into up to four concurrent tasks across conversations, including running multiple image generations at the same time while sharing the desktop safely.

• Delegate mode for bots: In Settings Tasks, you can let a bot delegate longer tasks to allowed fast, balanced, or strong models, automatically picking the account with the most usage left.

• Plan reviews on mobile: Read, comment on, edit, and approve or hand off plans directly from the mobile app, including building with Maestro or in a new conversation.

• Floating Computer window on remote workspaces: Pop out the Computer tab of a server workspace conversation over the chat or into a floating window with C / Ctrl+Alt+C.

• Broader model guidance and context limits: Custom context limits now compact conversations on Claude, GitHub Copilot, Grok, Cursor, and Google AI subscriptions, and Gemini 4 Argon / 3.8 Flash follow agent-tuned instructions across all supported providers.

If you run a bot server, back up your gateway volume before updating (the gateway database moves to schema 21).

Release notes: https://github.com/antonioducs/m...