Conductor coordinates four third-party coding agents across separate Git workspaces and a review workflow, while Meta already has Muse Code. Conductor Cloud is scheduled to shut down on November 20, leaving the future interface and its model neutrality unresolved.
Axios reported on November 4 that Meta acquired Conductor, founded in 2024 by Charlie Holtz and Jackson de Campos. Holtz said the company was joining Meta, and TechCrunch reported that Meta confirmed the acquisition. Consideration, transaction structure and a Meta product plan remain undisclosed.
What Conductor could add to Meta
Conductor's harness documentation lists Claude Code, Codex, Cursor and OpenCode. Its Git worktrees give independent coding tasks separate branches and working files; review tools bring diffs, checks and pull requests together for developers to evaluate. The external agents execute the coding tasks; Conductor organizes the workspaces and review flow.
Meta had already introduced Muse Code with persistent asynchronous subagents in August 2026. Conductor's possible incremental value therefore lies in its multi-provider interface and implementation experience, rather than the invention of parallel coding. Owning an interface could let Meta influence developer workflows. It would not automatically give Meta control over outside models or repositories.
Conductor Cloud's shutdown creates a distribution gap
On November 4, Conductor stopped new sign-ups and downloads. Its Cloud service is scheduled to become read-only on November 13 and shut down on November 20, when cloud-data deletion begins. The local product will receive new models and critical fixes until early 2027 before winding down. Conductor offers data export and workspace transfer, but has announced no Meta replacement or automatic migration.
Meta could use the team's expertise internally, integrate it with Muse Code or release another multi-provider product. The disclosed shutdown gives existing users a deadline. It does not tell them where their future work will move.
More agents also mean more bills and review
At Meta's published Muse Spark Standard rates, $1.25 per million input tokens and $4.25 per million output tokens, a hypothetical run consuming one million uncached input tokens and 200,000 output tokens costs $2.10 in inference. Five equivalent runs cost $10.50 before sandboxes, tests, retries or human review. These are arithmetic sensitivities, not observed Conductor costs or an estimate of the acquisition's return.
Software engineer Simon Willison describes the bottleneck: “The natural bottleneck on all of this is how fast I can review the results.” More parallel work creates value only if the benefit from accepted, reliable changes exceeds the combined cost of inference, execution, verification and rework. Conductor's product documentation provides no independent measurement establishing that net gain.
Competitors are pursuing the same workflow
OpenAI announced plans to acquire Ona for cloud agent execution in June 2026 and separately announced the proposed Pierre Computer acquisition for Git-linked code tools on November 4.
The Pierre agreement is subject to closing conditions and regulatory approvals. GitHub's agent management already brings competing agents into repository workflows.
The competitive issue is who becomes the place where tasks begin, code is reviewed and changes are approved; neither deal is a price comparable for Conductor.
What a successor would need to offer
A Muse-aligned interface looks more likely than a full recreation of Conductor's neutral cloud service, given Meta's existing tool and the announced shutdown. CTOL expects, with moderate confidence, that by November 5, 2027, Meta will not offer a generally available hosted successor supporting all four prior harnesses, separate Git workspaces and integrated pull-request review. Public product documentation and a working service accessible to new developers with all those capabilities would falsify that call.