Switch from Orchestrator to Task Machine
Orchestrator coordinates coding-agent sessions around worktrees and pull requests. Task Machine gives people and agents shared Tasks with plans, review gates and an Inbox for decisions.
Prefer the side-by-side comparison?Why people switch
-
Shared responsibility needs more
Orchestrator describes its local architecture as a single-user system built around a long-running daemon.
-
Local schedules need uptime
Automations run in the local daemon, so the computer must stay awake and Orchestrator must remain running. After a restart, it runs at most three missed occurrences.
-
Cloud work needs preservation
Orchestrator warns that a provider failure or missing checkpoint can leave Cloud work unavailable. It tells you to commit and push important work to the repository.
What maps to what
Every answer links to its source and shows when we last checked it. The oldest check on this page was November 15, 2026.
| Area | In Orchestrator | In Task Machine |
|---|---|---|
| Tasks |
In Orchestrator
A worker implements a task in its own workspace, an orchestrator supervises workers, and a separate reviewer checks a worker's pull request. |
In Task Machine
A person or an agent owns each Task. Its plan, discussion, decisions, runs and history stay together, so you can hand work between people and agents on the same record. |
| Recurring work |
In Orchestrator
Automations save a prompt, project, agent, role, timezone and schedule. You can inspect run history and pause or resume a definition. |
In Task Machine
Start a Task or Workflow once or on a repeat in your timezone. A recurring job keeps one stable root Task, with each occurrence's own status and history. |
| Approval |
In Orchestrator
Worker and orchestrator settings are independent. Available permission modes depend on the provider, and bypass-permissions lets provider tools act without their normal prompts. |
In Task Machine
Decisions land in your Inbox with context. Set each agent's actions to Ask first or Automatic, and use a Human approval step to hold a Workflow before publishing a report. |
| Instructions |
In Orchestrator
Project settings hold standing worker and orchestrator instructions. Rules files are read when a session starts, so edits apply to later sessions. |
In Task Machine
Keep run guidance in TAMA.md. Internal code review checks completed changes against enabled repository rules, and rule discovery proposes additions for you to review. |
| Code review |
In Orchestrator
Unresolved pull-request feedback returns to the owning worker. A reviewer runs separately from that worker and does not own its branch. |
In Task Machine
Agents address pull-request feedback and update the branch under your repository rules. Task Machine records its own approval of the exact commit before an agent may merge. |
| Execution |
In Orchestrator
Local projects use installed coding agents and their existing sign-ins. Cloud runs workers in remote workspaces and requires an enabled build and account access. |
In Task Machine
Run agents in Managed Cloud with nothing to install, or connect a Local Worker for files, tools and accounts on a computer you control. Local agents use Task Machine's tamacode runtime. |
| Pricing |
In Orchestrator
Local use is free. Cloud Starter is $20 a month and Pro is $40 a month on the website. Local agent usage is billed by the provider account or subscription each installed agent uses. |
In Task Machine
Usage based, from $29 a month before tax. The whole payment becomes usage that your workspace shares, with no base fee. |
A fact we cannot find or cannot check is left out, and prices can change after the date shown. Task Machine publishes this comparison and is an interested party. Orchestrator is named only to identify it, and there is no relationship between the two companies. Has something changed? Tell us.
You may stay with Orchestrator if
You may stay with Orchestrator if you want its free local app to coordinate the coding agents you already use. Its review loop routes unresolved comments back to the worker responsible. The companion phone app lets you answer agents, approve plans and follow sessions away from your desk. A working repository-centered setup may not need replacing.
What you give up
Task Machine's docs do not describe hosting your existing coding-agent terminals or a session-preserving importer for Orchestrator. Local agents use Task Machine's own tamacode runtime, so moving means rebuilding instructions and checking provider sign-ins rather than transferring a live session.
You also give up Orchestrator's free local orchestration. There is no trial for Task Machine. Budget for reconnecting repository access, recreating schedules and testing review boundaries before moving more work.
How the switch works
- Preserve unfinished work before cleanup. Orchestrator's Cloud guide says to commit and push important work. Its FAQ warns that deleting
~/.aopermanently deletes the database, session history and managed state. Keep a backup and check its export options for conversations and settings before uninstalling. - Create your workspace and connect the repository. Configure the Git provider's Connector separately for pull requests and reviews. Choose Managed Cloud or follow Local Worker setup for your own computer.
- Copy the selected job's prompt and relevant project rules into a new Task and repository guidance in
TAMA.md. Rebuild a repeated process in the Workflow builder, with a verifier for its output and an approval before an external action. - Start with Supervised autonomy. Test one small change, inspect its checks and delivered revision, and confirm that questions and approvals reach your Inbox before giving the agent more work.
- Recreate the schedule and timezone. Pause the original Automation before enabling the replacement for live work. Keep its history until you have preserved what you need, because deleting an Automation also deletes its run history.
Common questions
Is this the former Agent Orchestrator?
Yes. Its website identifies Orchestrator.inc as the new name for Agent Orchestrator and says it is the same app and ao command-line tool.
Can I bring my sessions and data?
Keep repository work in Git and preserve the local data described in Orchestrator's FAQ. Check its export options before removing anything. Task Machine's docs do not describe importing Orchestrator conversations or resuming its provider sessions. Recreate the job as a Task with the relevant instructions and files.
How do the costs compare?
Orchestrator is free locally, with model usage billed separately by providers. Its Cloud guide tells you to confirm account entitlements. Task Machine's pricing starts at $29 a month before tax, funds shared workspace usage and has no base fee or trial. Compare a representative job, including local fees and provider charges. You keep 100% of your revenue, Task Machine takes no cut and never custodies your accounts.
Can I keep using my installed coding agent?
You can keep it for separate work, but Task Machine does not host that terminal as a worker. Its Local Workers run tamacode, which can use supported subscriptions after you sign in. Test the instructions and results before replacing your existing process.