概览
What is Workflow Orchestrator MCP Server?
An AI-powered MCP server that manages and executes complex, dynamic workflows using a Large Language Model for intelligent decision-making. It is designed for developers who need adaptable, non-linear workflows within MCP-compatible clients like Cline.
How to use Workflow Orchestrator MCP Server?
Install dependencies with uv sync, set required environment variables (WORKFLOW_DEFINITIONS_DIR, WORKFLOW_DB_PATH, GEMINI_MODEL_NAME), then run uv run python -m orchestrator_mcp_server or the orchestrator-mcp-server command. The server exposes MCP tools: list_workflows, start_workflow, get_workflow_status, advance_workflow, resume_workflow. Workflows are defined as Markdown files in the configured definitions directory.
Key features of Workflow Orchestrator MCP Server?
- AI-driven decisions for dynamic branching and error handling.
- Workflows and steps defined in human-readable Markdown.
- Persistent state stored in a local SQLite database.
- Resumption of interrupted workflows with state reconciliation.
- Reusable, modular step definitions.
- Human-readable and editable workflow definitions.
Use cases of Workflow Orchestrator MCP Server?
- Automating multi-step code review and refactoring processes.
- Managing long-running task pipelines that can be paused and resumed.
- Creating adaptable GitLab issue analysis workflows.
- Generating commit suggestions or jokes with dynamic AI steps.
- Building custom orchestration logic for data processing or CI/CD.
FAQ from Workflow Orchestrator MCP Server
How is the server configured?
Configuration is done via environment variables: WORKFLOW_DEFINITIONS_DIR (path to Markdown workflow files), WORKFLOW_DB_PATH (path to SQLite database), GEMINI_MODEL_NAME (e.g., `gemini-2.
