Skip to main content
In Orkestral you don’t talk to a single assistant, you direct a team. Each agent has a role, a place in the hierarchy, and picks up the shared context of your workspace.

​
The hierarchy

  • You talk to the CEO in plain language.
  • The CEO reports to you, it reads the repositories, plans, and delegates.
  • The Tech Lead and Code Reviewer report to the CEO.
  • Specialists (Frontend, Backend, DevOps, QA, Designer…) report to the Tech Lead.
Every meaningful request becomes issues, the product is trackable work, not throwaway chat.

​
Roles

Your single point of contact. Reads the project, identifies the stack, breaks requests into issues and delegates to the right people. It can also propose and “hire” an initial team.
Coordinates the specialists, sequences the work and keeps the technical direction consistent.
Reviews code changes with senior-level rigor, structured findings, and a verdict (approved / changes requested) that comes back to you in chat.
Frontend, Backend, DevOps, QA, Designer and more, they execute the actual work in their area.

​
Hiring a team

Ask the CEO to assemble a team and it proposes an initial roster with a proper reporting structure. You approve, and the agents are created in your workspace.

​
Routing with @mentions

Mention an agent with @ to send a turn directly to them. That agent picks up the chat context and does the work itself, no extra delegation. 
@Code Reviewer check the auth changes in this PR
Use @ when you know exactly who should handle something. Leave it to the CEO when you want it planned and delegated for you.

​
The loop stays closed

When a specialist finishes, the work flows back: code-touching changes are routed to the Code Reviewer, and the verdict returns to your chat, so nothing is left “done but unverified.”