darnit

Generic compliance audit framework with plugin architecture.

Overview

darnit is the core framework that provides:

• Plugin System: Discover and load compliance implementations via Python entry points
• Configuration Management: .project.yaml for project metadata and file locations
• Progressive Verification: "Sieve" model for efficient compliance checking
• Remediation Framework: Auto-fix infrastructure for compliance gaps
• MCP Server Tools: Ready-to-use tools for AI assistant integration

Installation

pip install darnit

Usage

Discover Implementations

from darnit.core.discovery import get_implementation

# Get a compliance implementation by name
impl = get_implementation("openssf-baseline")
if impl:
    controls = impl.get_all_controls()

Configuration Management

from darnit.config.loader import load_project_config, save_project_config
from darnit.config.discovery import discover_files

# Discover existing documentation files
discovered = discover_files("/path/to/repo")

# Load project configuration
config = load_project_config("/path/to/repo")

Creating Implementations

To create a new compliance implementation:

1. Create a package with entry point darnit.implementations
2. Implement the ComplianceImplementation protocol
3. Register controls and checks

# pyproject.toml
[project.entry-points."darnit.implementations"]
my-standard = "my_package:register"

# my_package/__init__.py
def register():
    from .implementation import MyImplementation
    return MyImplementation()

See darnit-baseline for a complete example.

Package Structure

darnit/
├── core/           # Plugin system, models, discovery
├── config/         # Project configuration (.project.yaml)
├── sieve/          # Progressive verification pipeline
├── remediation/    # Auto-fix framework
├── server/         # MCP server tool implementations
└── tools/          # MCP tool helpers and utilities

License

Apache-2.0