MOFF CLI

Markdown Opinionated File Formatter

A command-line tool for validating and maintaining clean, organized documentation. Designed to work seamlessly with Large Language Models (LLMs) in modern IDEs like Cursor, VSCode, and Zed.

🎯 Purpose

MOFF helps maintain consistent documentation structure across projects by:

• Validating markdown files against configurable rules
• Enforcing location constraints for different document types
• Checking frontmatter schemas and required fields
• Ensuring headers follow specified patterns and order
• Providing visual feedback through tree visualization

Perfect for projects where documentation quality and consistency matter, especially when working with AI assistants that write markdown documentation.

✨ Features

• 📁 Smart Root Detection: Automatically finds documentation root via project_*.md files
• 🔍 Comprehensive Validation: Check frontmatter, headers, and file locations
• 🌳 Tree Visualization: See your documentation structure with error highlighting
• ⚙️ Configurable Rules: Define custom prefixes, patterns, and validation rules
• 💾 Result Persistence: Save validation results for CI/CD integration
• 🎨 Rich Terminal Output: Beautiful, colored output using Rich library
• 📚 Verbose Mode: Shows expected file structure templates when validation fails
• 🔧 Auto-fix: Automatically fix common issues like missing frontmatter and headers

📦 Installation

Using pip

pip install moff-cli

Using uv (recommended)

uv add moff-cli

🚀 Quick Start

1. Check your documentation (creates settings.json automatically if not present):

moff check

2. Visualize documentation structure:

moff tree

📖 Usage

Commands Overview

moff --help                    # Show help information
moff --version                 # Show version

moff check                     # Run validation checks
moff check --verbose          # Show expected structure for files with errors
moff check --fix              # Automatically fix fixable issues
moff check --save             # Run checks and save results to moff_results.txt
moff check --save --verbose   # Save results with expected structure templates
moff check --path ./docs      # Check specific directory

moff tree                      # Display documentation tree
moff tree --errors-only       # Show only files with errors
moff tree --no-check          # Skip validation (faster)

Example: Setting Up a Project

1. Create a project file (project_myapp.md):

---
project: myapp
---

# Overview

This is my application's main documentation.

## Requirements

- Python 3.12+
- Rich library

2. Create feature documentation (features/feature_auth.md):

---
project: myapp
feature: authentication
linked_features: ["users", "sessions"]
---

# Overview

Authentication system for the application.

## Requirements

- Secure password hashing
- JWT token support
- Session management

3. Validate your documentation (this will create settings.json if it doesn't exist):

moff check

Example Output

Check Command

Collecting documentation files...
Root directory: /Users/you/project/docs

✓ All checks passed!

No validation issues found.

Or with errors:

Validation Summary:
  Total issues: 2
  Errors: 2

Issues found:

features/feature_broken.md:
  error [feature] headers.missing: Missing required header level=2 text='Requirements' (line 10)

tech_database.md:
  error [tech] location.subdirs_only: File must be in a subdirectory, not in root

Or with verbose mode to see expected structure:

features/feature_broken.md:
  error [feature] frontmatter.missing: Required frontmatter is missing (line 1)
  error [feature] headers.missing: Missing required header level=1 text='Overview'
  error [feature] headers.missing: Missing required header level=2 text='Requirements'

  Expected structure for this file type (feature):
  ---
  project:
  feature:
  linked_features: []
  ---

  # Overview

  ## Requirements

Fix Command

When running moff check, you can use the --fix flag to automatically fix certain issues:

moff check --fix

Fixable issues include:

• Missing frontmatter blocks
• Missing required frontmatter fields
• Missing required headers

Example output:

Applying 3 automatic fixes...
Fixed features/feature_broken.md:
  • Added missing frontmatter block
  • Added missing header: Overview
  • Added missing header: Requirements

Fixes applied successfully!

Note: Some issues cannot be automatically fixed, such as:

• File location constraints (files in wrong directories)
• Header ordering issues
• Type mismatches in frontmatter values

Tree Command

📁 docs (documentation root)
├── 📁 features
│   ├── ⚡ feature_auth.md ✓
│   └── ⚡ feature_users.md ✓
├── 📁 technical
│   └── 🔧 tech_database.md ✓
└── 📋 project_myapp.md ✓

Summary:
  Total markdown files: 4
  Files with errors: 0
  Files with warnings: 0

✓ All files passed validation!

🐍 Programmatic Usage

When installed via pip, moff-cli can also be used as a Python library. Import it as moff_cli:

Basic Validation

from pathlib import Path
from moff_cli import Settings, Collector, Checker

# Load settings and collect documentation
settings = Settings()
collector = Collector(settings, start_path=Path.cwd())
collected_data = collector.collect()

# Run validation
checker = Checker(settings)
diagnostics = checker.check(collected_data)

# Process results
if not diagnostics:
    print("✓ All documentation is valid!")
else:
    for diag in diagnostics:
        print(f"{diag.path}: {diag.message}")

Custom Configuration

from moff_cli import (
    Settings,
    PrefixConfig,
    LocationConstraint,
    HeaderRule,
    HeaderMatch
)

# Create custom settings
settings = Settings()

# Add a custom prefix for API documentation
settings.prefixes["api"] = PrefixConfig(
    filename_pattern="api_*.md",
    location=LocationConstraint.SUBDIRS_ONLY,
    frontmatter_required={
        "endpoint": "string",
        "method": "string",
        "version": "string"
    },
    headers_required=[
        HeaderRule(level=1, text="Endpoint"),
        HeaderRule(level=2, text="Request"),
        HeaderRule(level=2, text="Response"),
    ]
)

# Save custom settings
settings.save_to_file(Path("settings.json"))

Tree Visualization

from moff_cli import Settings, Collector, Checker, TreeVisualizer
from rich.console import Console

console = Console()
settings = Settings()

# Collect and check
collector = Collector(settings)
collected_data = collector.collect()
checker = Checker(settings)
diagnostics = checker.check(collected_data)

# Display tree with error highlighting
visualizer = TreeVisualizer(settings, console)
visualizer.show_tree(collected_data, diagnostics)

CI/CD Integration

from pathlib import Path
from moff_cli import Settings, Collector, Checker, Severity

def validate_docs(project_path: Path) -> bool:
    """Validate documentation for CI/CD pipeline."""
    settings = Settings()
    collector = Collector(settings, start_path=project_path)
    checker = Checker(settings)

    # Collect and check
    collected = collector.collect()
    if collected.get("error"):
        print(f"ERROR: {collected['error']}")
        return False

    diagnostics = checker.check(collected)

    # Filter to only errors (ignore warnings)
    errors = [d for d in diagnostics if d.severity == Severity.ERROR]

    if errors:
        print(f"Validation failed with {len(errors)} errors")
        for error in errors:
            print(f"  ✗ {error.path}: {error.message}")
        return False

    return True

# Use in CI/CD
if not validate_docs(Path(".")):
    exit(1)

Available Classes and Functions

• Settings: Configuration management
• Collector: File discovery and parsing
• Checker: Validation engine
• TreeVisualizer: Tree visualization
• Diagnostic: Validation issue representation
• Severity: Error, Warning, Info levels
• LocationConstraint: ROOT_ONLY, SUBDIRS_ONLY, ANY
• HeaderOrder: STRICT, IN_ORDER, ANY
• HeaderMatch: EXACT, REGEX

⚙️ Configuration

MOFF uses settings.json for configuration. The default configuration supports three document prefixes:

Default Prefixes

Prefix	Pattern	Location	Purpose
project	project_*.md	Root only	Main project documentation
feature	feature_*.md	Any	Feature specifications
tech	tech_*.md	Subdirs only	Technical implementation details

Custom Configuration Example

{
  "version": 1,
  "root": {
    "detect": {
      "method": "project_file",
      "pattern": "project_*.md"
    },
    "override_path": null,
    "ignore": [
      "**/.git/**",
      "**/.venv/**",
      "**/node_modules/**",
      "**/archive/**"
    ]
  },
  "prefixes": {
    "api": {
      "filename": {
        "pattern": "api_*.md"
      },
      "location": "subdirs_only",
      "frontmatter": {
        "required": {
          "project": "string",
          "endpoint": "string",
          "method": "string"
        },
        "optional": {
          "deprecated": "boolean"
        }
      },
      "headers": {
        "required": [
          {
            "level": 1,
            "text": "Endpoint",
            "match": "exact"
          },
          {
            "level": 2,
            "text": "Request",
            "match": "exact"
          },
          {
            "level": 2,
            "text": "Response",
            "match": "exact"
          }
        ],
        "optional": [],
        "order": "in-order"
      }
    }
  }
}

Configuration Options

Root Detection

• detect.method: Currently supports "project_file"
• detect.pattern: Glob pattern for root detection (default: "project_*.md")
• override_path: Bypass auto-detection with explicit path
• ignore: List of glob patterns to exclude

Location Constraints

• "root_only": File must be in root directory
• "subdirs_only": File must be in a subdirectory
• "any": File can be anywhere

Frontmatter Types

• "string": Text values
• "number": Numeric values (int or float)
• "boolean": True/false values
• "list": Array values
• "object": Dictionary/object values

Header Order

• "strict": Headers must appear in exact order
• "in-order": Headers must be in order but others can appear between
• "any": No order enforcement

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Built with Rich for beautiful terminal output
• Uses markdown-to-data for parsing
• Inspired by the need for better documentation tooling in AI-assisted development