... last release update I posted here was 0.17.9. We're now at 0.17.14, with a few additions and a lot of work on the small things you feel while editing. Here's what's changed since then: OpenCode joins the Agents panel, alongside Claude Code, Codex, and Grok Build. Both OpenCode

... Version, command and log details are available under advanced options. If the available version cannot be queried, the UI offers retry and network settings instead of claiming an update is available. The setup panel covers Claude Code, Codex CLI, OpenCode, Pi Agent, Grok Build and Antigravity. Pi is new in this release, including native session restore, task execution, status events and local MCP integration. Thanks to wuyh410 for PR #33. For an installed Antigravity CLI, upgrades now use its official ... ... workbench layout, not the new setup panel. For people using several agent terminals: would you prefer completion reminders to stay until explicit acknowledgement, or disappear when you focus the relevant terminal? The tradeoff here is avoiding missed work versus

... explain everything to the next agent, lose all the context. felt dumb paying for 3 plans and still getting blocked. so i built codemax. one terminal, all your AI coding agents running side by side (claude code, codex, opencode, grok, cursor). when one hits its limit, codemax hands the task + full context to the next one. u just keep working. - no api keys. uses the subscriptions u already pay for - free + open source (AGPL) - gemini cli + copilot coming next ...

... much of the week and the 5-hour session is left - when each one resets - whether you're on pace, or will run out before the reset It works with Claude, Codex, GitHub Copilot, Cursor, Antigravity, Grok, Ollama, and OpenCode, for every account signed in on the Mac, including several Claude Code profiles. The Agents tab also lists every coding agent running, what each one started (dev servers, tests, browsers), and the tokens Claude Code and Codex used today ... ... smaller chunks so a cutoff doesn't land mid-edit. I also commit before every long agent run, so hitting a limit just means "pick up from the last commit" instead of untangling half-finished changes. The "on pace vs

... about a small interaction with a useful consequence: ask one coding agent to send work to another terminal, then read back the result. AI-assisted capability diagram, not a screenshot. With local MCP enabled, new Claude Code, Codex, OpenCode, Grok Build and Antigravity terminals launched by Termexo can connect automatically. A connected agent can find an existing terminal, send a prompt, submit it and read its output. The task board can also launch configured Claude Code, Codex, OpenCode

... yesterday and ended the day at #1. 327 upvotes and a wall of thoughtful comments. For a bootstrapped team, that caught us off guard, especially in a field this strong. We launched the same day as some great teams, @opencode , @Fina @Kai , and the folks behind them, @thdxr , @clay_raterman , @jim_hartung . Good day for all of us Big thanks to @chrismessina for hunting us again. The pre-launch calls shaped the whole thing, the story, the flow, the maker ...

... maintain Termexo, an MIT-licensed Windows workbench for coding agents. Version 0.10.10 adds 19 local MCP tools for terminals, tasks and selected settings. New Claude Code, Codex, OpenCode, Grok Build and Antigravity terminals launched in Termexo can automatically load the connection after you enable and apply MCP settings. Existing agent processes need to restart, and the desktop must stay open. A practical first trial: ask termexo-desktop to list workspaces, with no terminal input or settings changes. Then read ... ... loopback. Terminal, task and settings permissions are separate; settings access needs enabling. The stdio bridge reads the current credential for each request rather than putting the token in generated configuration. Automatic configuration has different scope: Claude Code, Codex and OpenCode

... practical question for people running several coding agents: which terminal needs attention, and what can you trust when you come back to it? Agent status now follows the actual lifecycle. In v0.10.8 I checked Claude Code, Codex, OpenCode, Grok Build, and Antigravity against their installed CLI behavior. A late startup or idle event no longer overwrites a newer running state. OpenCode

... since I posted here hehe. But I m back with something we ve been building for a while. Tomorrow, Murmell Bots are coming. Your AI bots, in the cloud, ready to work for you. Claude Code, Codex, Gemini, Kimi, OpenCode, all in one place. And this is just the beginning. See you tomorrow ...

... Maintainer update: Termexo v0.9.0 is out for Windows 10/11 x64. Antigravity joins Claude Code, Codex CLI and OpenCode, with local conversation resume, model and reasoning controls, /usage allowance and live activity status. This release also adds vendor Windows installer choices for supported CLIs, a direct install entry when an agent is missing, change-block diff navigation (Alt+Up/Down), and fixes long lines crossing split-diff columns. OpenCode

... Termexo v0.10.4 is live. Grok Build joins Claude Code, Codex, OpenCode, and Antigravity in our Windows terminal workbench. With Grok Build CLI installed and signed in, you can launch it in a workspace, list and resume sessions, and use it on the task board. The usage panel now recognizes an exhausted free allowance, while an unknown percentage stays unavailable instead of appearing as 0%. This release also fixes a frustrating long-session issue: the session change count could fall ...

... kept re-checking their work by hand, so I built foolscap to do the checking for me. How it works: you give it a task. GPT-6 Astra plans it and hands the work to Claude Code, Codex or OpenCode running on your machine. Then foolscap reads what the test command actually printed and decides whether the run passed. A model doesn't make that call. Two rules do most of the work: Repairs are capped (two for implementation ...

... Maker update: Termexo v0.8.5 is out. I maintain this MIT-licensed Windows workbench for Claude Code, Codex CLI, OpenCode and real terminals. More consistent terminal behavior The EXE, MSI and npm packages now bundle Microsoft's MIT-licensed ConPTY components. Older Windows ConPTY builds can lose Shift+Tab or mouse reports, affecting Claude Code shortcuts and OpenCode

... Existing workbench screenshot (v0.8.0), shown for context. I'm the maintainer of Termexo, an MIT-licensed Windows workspace for Claude Code, Codex and OpenCode. v0.8.7 focuses on a frustrating reliability problem: an agent could keep running while its terminal went blank and appeared stopped. Resizing around a wide character could trigger a parser panic, poison the screen mutex and leave subsequent screen reads failing. Switching between clients could reproduce that combination. The screen model now uses avt, the terminal model ...

... Termexo v0.8.6 AI-generated release illustration I maintain Termexo, an MIT-licensed Windows workbench for Claude Code, Codex and OpenCode. Version 0.8.6 focuses on reloading a view, reconnecting remotely, resizing a window and switching to a terminal that was hidden. Keep the running process Previously, frontend loading could be mistaken for a new terminal launch. Reloading could restart an agent that was already working. The view now reconnects to a backend terminal that is still alive, preserving its current launch ...

... Maker update: Termexo v0.8.3 is out. I maintain this MIT-licensed Windows workbench for Claude Code, Codex CLI, OpenCode and real terminals. If typing has felt delayed while agents are working, this is an update worth installing. Give keystrokes room to run Git queries and event synchronization previously queued on the same IPC thread as terminal input. The release moves background commands to a thread pool. The official release reports key-input latency p99 falling from ...

... manually, or the auto-updater will do the job for you in a few hours :) (FYI: The auto-updater polls every 6 hours, and it has been made opt-in) This update carries the hottest requested features such as OpenCode support, tracking subagents and fixing the shocked state of the cat which I hope will make more sense now. The Changelog is as follows: Comment from Hossein Yazdi(@hosseinyazdi): OpenCode

... servers, so you can just pick up right from where you left off, but now you can close your laptop lid! One of the things we're most curious about is why people are still using local agents like OpenCode or Claude Code. Is it stickiness? Lack of good cloud solutions? Cost? Or something else? Would love to know! ( p.s, we have a super ultra rare promotion running right now, so you can try out the full platform for free ... ... tools are in scope, keep latency predictable, and avoid the "works in cloud, breaks on my machine" drift. If cloud hand-off can preserve that determinism and make approval plus secret exposure obvious at the transfer point, the local-vs

... Product Hunt. Murmell already runs Claude Code, Codex, Kimi, and OpenCode on the same cloud canvas. More agents are coming, and I want to prioritize what you'd actually use. If you had to pick the next one we should wire in first, what would it be? 1. Hermes 2. OpenClaw 3. Something else (tell me which) Drop your vote below, and if it's "something else," name the agent. I'll build toward whatever gets the most pull :) Comment ...

... Claude Code, Antigravity, and OpenCode, three different companies, same repo, same time. They discussed. Nothing broke. All free, all running locally on my machine. Bothread is a local room your coding agents join together , so instead of quietly overwriting each other's work, they coordinate. The big thing: the agents talk to each other while they work. One tells the others "I changed this interface, adjust for it," and it actually lands. Built solo. Free, open source , runs entirely ...

... work, or whether your cache paid off? LangWatch now tracks every Claude Code session: tokens, costs, cache hits, bash commands, MCP tool calls, file edits, and full terminal replays. Works with Claude Code, Codex, Gemini CLI, and opencode. If you want to support us, feel free to do so You can find us under "LangWatch" on the launching page! Comment from Manouk Draisma(@manouk_dr): Comment from Manouk Draisma(@manouk ...

... worker. For UI testing, a remote headed browser is not just a convenience. It changes the debugging loop. Instead of asking "what did the script think happened?", you can ask "what did the user actually see?" ### Harness Code vs Local Agent Harness Another important decision is where the agent capability should come from. One option is to build the entire reasoning layer yourself: call an LLM API directly, manage prompts, retries, tool calls, screenshots, JSON parsing, and safety rules. That ... ... gives maximum control, but it also means the test harness becomes an AI platform project. The alternative

... knowledge-work domain now has a decent tool. Harvey and Legora for legal. Rogo for finance. All decent and all narrow and not accessible to everyone. Coding has something the rest of knowledge work doesn't: harnesses. Cursor CLI, OpenCode, Codex, Claude Code general-purpose agents that turn a model into a real force multiplier. So where's the equivalent for everyone else? Why is there no publicly available harness that's a genuine upgrade over Claude Code or Codex ...

... automatic maintenance and compaction BM25-powered memory recall with citations an MCP server so the project memory is also available from Claude Desktop and other MCP clients support for additional coding agents such as Codex CLI, Gemini CLI and OpenCode numerous bug fixes, security improvements and CI enhancements The goal wasn't to reinvent OpenWolf, but to make it something I could comfortably use every day on long-running projects without worrying about maintenance or ever-growing metadata ...

... Update available - v0.7.0 New: We have added support for OpenCode and few fixes to claude cli version To update, run both commands: npm i -g idledev-cli@latest idledev install Then restart Claude CLI/OpenCode ...