Documentation navigation
Getting started Quickstart Installation Core concepts Product tour
Building trees Creating nodes Prompt blocks Connectors Roles and governance Skills Co-pilot Collaboration
Run and inspect Execution Runs and traces Versions and rollback
SDK Python SDK Node.js SDK
Integrate API MCP and bundles
Providers AI connections Hosted providers for testing
Deploy Self-hosting Orchetree Cloud Environment variables

Product tour

Orchetree separates system design, execution, and infrastructure settings so each workflow has a clear home.

Projects

The Projects page is the entry point for a workspace. A project groups its agent trees, reusable connectors, reviewed skills, versions, and run history.

Open a project to use:

  • Trees for the multi-tree map and individual tree editor
  • Connectors for reusable API, database, retrieval, search, and MCP tools
  • Skills for security-reviewed instructions that can be attached to nodes
  • Runs for execution, traces, events, output, cost, and latency

The Project Trees Map shows delegation relationships between trees and lets you move from the system-level architecture into one tree.

Tree editor

The tree editor is where developers and business experts work on the same system within different ownership boundaries.

Select a node to edit its capability, prompt blocks, execution policy, model connection, connectors, and skills. The graph records explicit parent-child relationships; the backend planner derives execution order from those relationships.

See Nodes, Prompt blocks, and Roles and governance.

Showcase Project

The built-in Showcase Project is a working reference implementation authored with the Orchetree SDK and imported through the normal bundle path. Its Specialist Trees power the embedded Co-pilot and demonstrate orchestration, prompt coaching, agent authoring, workspace analysis, testing, and proposed actions.

Its executions appear in the same Runs area as user-created systems, so the reference remains inspectable rather than being a hidden demo.

Settings

Use Settings > AI Connections to add, test, edit, and select model providers. A running tree needs a resolvable AI connection.

See AI connections.

Run and inspect

From a tree or the Runs area, provide JSON input and start an execution. The run record tracks status, output, node traces, events, tokens, latency, and errors. Runs can be cancelled, repeated, exported, and analysed.

See Execution and Runs and traces.

Code-first and integrations

The Python and Node.js SDKs define and synchronize project structure from source code. GraphQL, REST, WebSocket subscriptions, MCP, and portable bundles cover external integration and deployment workflows.

See Python SDK, Node.js SDK, API, and MCP and bundles.