Superconductor is a multiplayer AI workspace for your team and coding agents. 👥🤝🎮 Multiplayer · Your teammates can join your agent sessions (incredible for code review) · Manage dev env, integrations, and token spend for the whole team 🤖🧠⚡ Multi-agent · Claude, Codex, Pi, etc all in one place (and you can use your subscriptions) · Launch multiple agents in parallel, select best one with automatic QA Checks · Chief of Staff agent that tracks all work and can message any agent or human on your team 📱💻☁️ Accessible anywhere · Web, desktop, iPhone have 100% feature parity · Or, use from Slack, GitHub, or even via email · Or -- and this is crazy -- add it to a live meeting and see prototypes in real time as you discuss feature ideas Free to get started, bring your own keys/subscriptions. superconductor.com

New in Superconductor last week: custom MCP connectors, Sentry OAuth, Opus 5.5, GPT-6 Sol/Luna, Grok 4.7, smoother diff reviews, and more. Full changelog: superconductor.com/blog/week…

What does it mean to work with AI agents in a fundamentally multiplayer way? I can review a pull request by chatting with the EXACT SAME agent that was used to develop it (which cost almost $700 lol). I can see the code diff right there, and leave comments to the agent on it. I can also easily pull in the human developer with an @-mention. And I never have to check code out locally. The live app preview is running on the agent's sandbox, so I can use it if I don't trust the agent's (or the QA agent's) screenshots/videos of the feature.

When multiple people are collaborating with coding agents in the same chat, it's important for the agents to know who is who. Agents on Superconductor now distinguish between team members and can summarize who contributed what to the conversation. superconductor.com/blog/mult…

that feeling when you can't use your best buddy @Superconductor to talk through the maintenance operation for Superconductor.com 😭

Never copy and paste.

Want to chat with the agent without changing code? Check out the new Chat tab in Superconductor. Ask agents what got merged today, to draft a contract, or to scope out a feature. Or, launch multiple agents on the same question and compare replies. superconductor.com/blog/chat…

Meet Sergey Karayev (@sergeykarayev), he's building @Superconductor, the multiplayer AI platform for teams and agents. Sergey has been writing Ruby for over 20 years and will be sharing his learnings from making AI tools go from single-player to multiplayer. Don't miss his talk: luma.com/sfrubyconf2026

New in Superconductor this week: Gemini 3.8 Flash, DeepSeek V4.1 Flash, per-chat network access, more readable work logs, and more. Check out the full weekly changelog here: superconductor.com/blog/week…

For over a year, my team has been working with AI agents in a fundamentally different way than most. All of our Claude/Codex/Pi/etc agents run in the cloud, and any session is joinable by anyone on the team. Each of our meetings has an agent that launches subagents to do research, draft posts, and implement features as we discuss things. A Chief of Staff agent lets me know what's waiting for my review, and can talk to any person or agent on the team to resolve bottlenecks. Working in this fundamentally multiplayer way has given us a preview of the way everyone will work soon, so I wrote up a short manifesto explaining • the current problems • principles for a great solution • some things that are tricky to get right Check it out, and let me know what you think! multiplayer-ai.com

Agreed. Adding an agent to a group chat is a start, but not the end. We've been thinking about multiplayer AI for a while and wrote up five principles that a great solution should have. 1. Never copy-and-paste Most work doesn't happen in chat, but in docs, emails, Salesforce, GitHub, etc. The agent has to live next to the work, and everyone involved in the work has to be able to talk to the same agent, instead of endlessly copying and pasting back and forth. 2. Work with the door open When chat sessions are private, good prompts and good habits spread slowly. When they're shared, everyone gets better fast. Shopify's internal agent is only reachable from public Slack channels (@tobi called it "learning on the shop floor".) 3. Continuously improve When someone writes a prompt that nails it on the first try, the rest of the team should learn from it. When the agent needs correcting, that correction should turn into a reusable skill. When the same workflow runs over and over, its quality should start being measured automatically. 4. People are not routers A human should never be asked a question the agent already has the answer to. Chasing status updates and relaying answers between people is work for an agent. 5. Nothing starts from scratch When sessions live in the cloud, you pick up every project exactly where you left off. Every doc, plan, and PR should have a session behind it you can resume months later. A new hire is productive on day one. There are a few more considerations, on security and build-vs-buy, on the site. Curious to get your take! multiplayer-ai.com

it feels surprisingly good to manage a fleet of implementations via chief of staff lol

Which coding agent wins on YOUR codebase? Benchmark agents against your team’s merged PRs and choose the one that fits your repo and stack.

Excellent set of results, matching ours! You can build your own "Custom SWE-Bench" on any repo -- we built one for our RoR codebase, and the results match these: • Kimi K3 is just as good as Opus/Fable 5! • GPT 5.6 is cheap but worse • GLM 5.2 worse yet

💪 Grok 4.6 joins the Pantheon of our benchmark, together with Opus/Fable 5 and Kimi K3. 💪 A good and fast model, and my new daily driver. Check out the video for more details, an explanation of how our benchmark works, and how you can make your own!

You can now just forward an email to an agent and get an email back when done. Here I have GPT 5.6 redline a contract, which it emails me back when done. Hard to go back from this workflow!

Same app, now officially approved and available in the @SlackHQ Marketplace. Summon Superconductor from Slack to work on a task, and give your agents the entire thread as context for their work. slack.com/marketplace/A09AY6…

Coding agents should not grade their own homework. Our coding agent said it was done, with tests passing. Then our QA agent opened the app and found the feature was broken. Possibly my favorite thing we added recently!

Coding agents belong in remote sandboxes. When they run locally, they often sit next to a developer’s laptop access: SSH keys, cloud credentials, package tokens, browser sessions, VPN routes, local files, internal systems, and years of accumulated setup. That may be fine for individual experimentation. For companies standardizing agent usage across engineering teams, “where does the agent run?” becomes a security architecture question. Remote sandboxes don’t make agents magically safe. You still need tool-level controls, scoped permissions, and human review. But they let the company define the boundary: - which secrets the agent receives - which network paths it can reach - which logs are retained Local sandboxing is useful, and teams should use it. But local and remote sandboxes solve different problems. Local sandboxing adds controls inside one developer-controlled workstation. Remote sandboxing gives the organization a team-controlled boundary around agent work. Full writeup here: superconductor.com/blog/remo…

Kimi K3 is the best open weight model on our custom SWE-bench. On our Rails codebase it lands right at Opus 4.8 quality for a fraction of the cost. Next open weight model down, GLM 5.2, is about 8 pts lower! It's available on @Superconductor, via Pi & OpenRouter via Fireworks.