flowmason 1.0.18

Universal AI Workflow Infrastructure - Build, debug, and deploy intelligent pipelines

FlowMason

Universal AI Workflow Infrastructure - Build, debug, and deploy intelligent pipelines.

Overview

FlowMason is an AI pipeline orchestration platform that enables developers to design, build, debug, and deploy intelligent workflows. It uses a Salesforce DX-style hybrid model:

• Development: File-based pipelines (.pipeline.json) in VSCode with Git version control
• Deployment: Push to staging/production orgs where pipelines run from databases
• Runtime: Backend APIs expose pipelines for consumption

Features

• Visual Pipeline Builder - Design AI workflows visually in VSCode or Studio
• Three Component Types - Nodes (AI), Operators (deterministic), Control Flow
• Full Debugging - Breakpoints, step-through, prompt iteration
• Package System - Distribute components as .fmpkg files
• Multi-Environment - Local development, staging, production orgs
• Enterprise Ready - API keys, RBAC, audit logging, SSO/SAML

Installation

Requirements

• Python 3.11 or higher
• pip (Python package manager)

Install from PyPI

pip install flowmason

What Gets Installed

When you install FlowMason, the following CLI commands are automatically added to your PATH:

Command	Description
fm	Short alias for FlowMason CLI
flowmason	Full FlowMason CLI command

Both commands are identical - use whichever you prefer.

Verify Installation

# Check version
fm --version
# or
flowmason --version

# See all available commands
fm --help

Troubleshooting

If fm or flowmason commands are not found after installation:

1. Check Python scripts directory is in PATH:

   # Find where pip installs scripts
   python -m site --user-base
   # Add the bin directory to your PATH (add to ~/.bashrc or ~/.zshrc)
   export PATH="$PATH:$(python -m site --user-base)/bin"
   

2. Try using Python module directly:

   python -m flowmason_core.cli.main --help
   

3. Reinstall with --user flag:

   pip install --user flowmason
   

4. Use a virtual environment (recommended):

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate
   pip install flowmason
   fm --version
   

Quick Start

# Initialize a new project
fm init my-project
cd my-project

# Start the local Studio backend
fm studio start

# Run a pipeline
fm run pipelines/main.pipeline.json

# Debug a pipeline
fm run pipelines/main.pipeline.json --debug

VSCode Extension

Install the FlowMason extension from the VS Code Marketplace for:

• IntelliSense for decorators and component patterns
• Visual DAG editor for pipelines
• Debugging with breakpoints and step-through
• Test Explorer integration
• Prompt iteration during debug

Define Components

Node (AI-powered)

from flowmason_core import node, NodeInput, NodeOutput, Field

@node(
    name="summarizer",
    category="reasoning",
    description="Summarize text using AI",
)
class SummarizerNode:
    class Input(NodeInput):
        text: str = Field(description="Text to summarize")
        max_length: int = Field(default=100)

    class Output(NodeOutput):
        summary: str

    async def execute(self, input: Input, context) -> Output:
        provider = context.providers["anthropic"]
        response = await provider.call(
            prompt=f"Summarize in {input.max_length} words: {input.text}"
        )
        return self.Output(summary=response.text)

Operator (Deterministic)

from flowmason_core import operator, OperatorInput, OperatorOutput

@operator(
    name="json-transform",
    category="transform",
    description="Transform JSON data",
)
class JsonTransformOperator:
    class Input(OperatorInput):
        data: dict
        expression: str

    class Output(OperatorOutput):
        result: dict

    async def execute(self, input: Input, context) -> Output:
        import jmespath
        result = jmespath.search(input.expression, input.data)
        return self.Output(result=result)

CLI Commands

# Pipeline execution
fm run <pipeline.json>              # Run pipeline from file
fm run --debug <pipeline.json>      # Run with debugging

# Project management
fm init                             # Initialize project
fm validate                         # Validate pipelines

# Studio management
fm studio start                     # Start local Studio
fm studio stop                      # Stop Studio
fm studio status                    # Check status

# Org management (staging/production)
fm org login --alias staging        # Login to org
fm deploy --target staging          # Deploy pipelines
fm pull --target staging            # Pull pipelines

# Package management
fm pack                             # Build .fmpkg
fm install <package.fmpkg>          # Install package

Documentation

Visit https://flowmason.com/docs for complete documentation.

Support

• Documentation: https://flowmason.com/docs
• Support: https://flowmason.com/support
• Email: [e-mail]

License

Proprietary. See LICENSE file for terms.

Copyright (c) 2025 FlowMason. All rights reserved.