Meta Bought Conductor and Its Cloud Goes Read-Only November 13. What a Nine-Day Warning Teaches About Building on Small AI Tools.
On November 4, Conductor announced it is joining Meta. If you used its cloud workspaces to run coding agents in parallel, you have until Friday, November 13 before chats become read-only, and until Friday, November 20 before Conductor Cloud shuts down and deletion of cloud user data begins. That is nine days from the announcement to read-only, and one week from today.
The timeline
Everything below comes from Conductor's own announcement, and TechCrunch's report agrees on the November 20 cloud shutdown. As of November 4, new users cannot sign up or download the app, and free users cannot upgrade. Pro users keep pro features for free. On November 13, Conductor Cloud goes read-only: you can read old chats and export data but not send new messages, and active subscribers are refunded. On November 20, the cloud shuts down and deletion of cloud user data begins.
The local app is on a different clock. Conductor says it keeps working as normal, with new models and critical bug fixes, until early 2027, and then winds down. The FAQ says all user data will be deleted when the app shuts down in early 2027. The founder says something new is coming early next year.
Meta confirmed the acquisition to TechCrunch but gave no price or terms. Neither source I could read states how many users Conductor had or how much it raised, so I will not guess.
How to get your work out
The FAQ gives two paths. To move a single cloud workspace and its chats to a Mac, right-click the workspace and choose "Transfer to local". To download everything, go to Settings, then Account, then Export data, which produces an archive. I would do both for anything that matters, and I would do it this week, not on the 12th. Read-only mode on the 13th is the point where a failed export stops being fixable by re-sending a message.
The lesson I keep relearning
Conductor's post also says the company has never sold or trained on user data and will not start now. I take that at face value, and it still does not change the practical problem: the tool I built a workflow around now has a calendar. Small AI developer tools get absorbed by bigger companies often enough that the pattern is worth planning for. Here signups closed the same day, the paid tier became free, and the clock started.
The risk is not that the product gets worse. It is that your agent history, your workspace layout and your habits live in somebody else's format, on somebody else's schedule. For a solo operator, that history is the closest thing you have to institutional memory.
What I'd actually do
I would keep the output of agents in git, not in the tool. Branches, commits and a short notes file in the repo survive any acquisition. I would also keep a one-line entry for each AI tool I depend on: what it holds, how to export it, and what I would switch to. When Conductor's notice landed, anyone with that list had a ten-minute task. Anyone without one had a project.
For Conductor specifically, if you are a user: export now, transfer active workspaces to local if you want to keep working, and decide whether to stay on the local app until early 2027 or move to another parallel-agent setup while the cost of switching is low.
Where I could be wrong
The local app may stay good enough through early 2027, and the "something new" from Meta may be better than what you lose. Acquisitions sometimes work out for users. I would still treat the export as free insurance, because the downside of skipping it is permanent and the cost of doing it is a few minutes.
Author
Lukas
@lukcombinator