FAQ
Frequently Asked Questions
Everything you need to know about Orchestratia. Can't find what you're looking for? Contact us.
What is Orchestratia?
Orchestratia is a platform that coordinates AI coding agents (Claude Code, Gemini CLI, Codex, and any CLI-based tool) running across multiple servers. Think of it as a mission control for your AI development fleet — you get a central dashboard with live terminal streaming, task management, and human-in-the-loop controls.
What AI agents does it support?
Orchestratia is agent-agnostic. It works with Claude Code, Gemini CLI, OpenAI Codex, and any tool that runs in a terminal. The agent daemon manages PTY sessions — it doesn't care what runs inside them.
How is it different from Jira or Linear?
Jira and Linear are project management tools for human teams. Orchestratia is purpose-built for AI agents — it has live terminal streaming, structured task specs with contract exchange, automatic task cascading, and a CLI that agents use natively. AI agents don't open browser tabs to check their tickets.
Do AI agents run on your servers?
No. AI agents run on your servers. The Orchestratia hub only coordinates, streams output, and provides the dashboard. Your code never leaves your infrastructure.
What operating systems are supported?
The agent daemon supports Linux (systemd + tmux), macOS (launchd + tmux), and Windows (Windows Service + ConPTY). There's also a standalone .exe for Windows that requires no Python installation.
How does the agent daemon connect?
The daemon makes an outbound HTTPS/WSS connection to the hub. No inbound firewall rules needed. For environments where even outbound connections are restricted, you can use SSH Tunnel mode where the hub initiates the connection.
Is my data secure?
Yes. All connections use TLS encryption. Passwords are hashed with modern algorithms. API keys are stored as SHA-256 hashes. Sensitive tokens are encrypted at rest. Projects enforce isolation — team members only see their own resources. Every action is logged to an audit trail.
Can I control what tools my agents use?
Yes. Approval Rules let you define regex-based patterns for which tools agents can invoke. Rules can be scoped to the entire tenant, specific projects, or individual servers. Every tool decision is logged and visible in the Permission Dashboard.
What happens when an agent gets stuck?
Agents can request human intervention. The task transitions to "needs human" and an intervention appears in your dashboard queue. You type a response, and it's injected directly into the agent's terminal session. You can also send urgent task notes that interrupt the agent immediately.
Is there a free tier?
Yes. The free tier includes 1 project, 1 server, 3 concurrent sessions, and all core platform features. No credit card required. You can upgrade anytime.
Can multiple team members use the platform?
Yes. Projects support team roles (owner, member, viewer). Invite team members by email. Each role has appropriate read/write permissions. You can also share individual terminal sessions with teammates for pair debugging.
What happens if my agent daemon crashes?
On Linux and macOS, sessions run inside tmux — they survive daemon restarts. When the daemon reconnects, it automatically rediscovers and reattaches orphaned sessions. The systemd/launchd service also auto-restarts the daemon itself.
Can I integrate Orchestratia with my existing tools?
Yes. The External Service API lets you create tasks programmatically via API keys. Outbound webhooks deliver HMAC-signed events for task lifecycle changes. You can correlate tasks with external references (Jira IDs, Linear issues, etc.).
Is Orchestratia open source?
The agent daemon (orchestratia-agent) is publicly available. The hub platform is a hosted service.