debtmap

Beta Software - Debtmap is actively developed and tested in production. Core features are stable for Rust, though APIs may evolve as we add new capabilities. Contributions and feedback welcome!

Debtmap combines coverage-risk correlation with multi-factor analysis (complexity, dependencies, call graphs) and entropy-adjusted scoring to reduce false positives and prioritize testing efforts effectively.

📚 Read the full documentation for detailed guides, examples, and API reference.

Why Debtmap?

Debtmap answers two critical questions:

"What should I refactor to reduce cognitive burden?" - Identifies overly complex code that slows down development
"What should I test first to reduce the most risk?" - Pinpoints untested complex code that threatens stability

Unique Capabilities:

Coverage-Risk Correlation - Combines complexity metrics with test coverage to identify genuinely risky code (high complexity + low coverage = critical risk)
Reduced False Positives - Uses entropy analysis and pattern detection to distinguish genuinely complex code from repetitive patterns, reducing false positives by up to 70%
Actionable Recommendations - Provides specific guidance with quantified impact metrics instead of generic warnings
Multi-Factor Analysis - Analyzes complexity, coverage, dependencies, and call graphs for comprehensive prioritization
Fast & Open Source - Written in Rust for 10-100x faster analysis, MIT licensed with no enterprise pricing

📖 Read more: Why Debtmap?

Documentation

📚 Full Documentation - Complete guides, tutorials, and API reference

Quick Links

Getting Started - Installation and first analysis
CLI Reference - Complete command documentation
Configuration - Customize thresholds and behavior
Analysis Guide - Understanding metrics and scoring
Coverage & Risk - Integrate test coverage data
Examples - Common workflows and use cases

Quick Start (3 Minutes)

Install

curl -sSL https://raw.githubusercontent.com/iepathos/debtmap/master/install.sh | bash

# For test coverage analysis (optional)
cargo install cargo-llvm-cov

Analyze

# Basic analysis
debtmap analyze .

# With test coverage (recommended)
cargo llvm-cov --lcov --output-path target/coverage/lcov.info
debtmap analyze . --lcov target/coverage/lcov.info

# Generate JSON report
debtmap analyze . --format json --output report.json

# Generate interactive HTML dashboard
debtmap analyze . --format html > report.html

Review Results

Debtmap shows you exactly what to fix first with actionable recommendations:

#1 SCORE: 8.9 [CRITICAL]
├─ TEST GAP: ./src/parser.rs:38 parse_complex_input()
├─ ACTION: Add 6 unit tests for full coverage
├─ IMPACT: -3.7 risk reduction
└─ WHY: Complex logic (cyclomatic=6) with 0% test coverage

Concise Actionable Recommendations

Debtmap provides step-by-step recommendations with clear impact estimates and difficulty levels. Each recommendation includes:

Maximum 5 high-level steps - Focused, actionable tasks
Impact estimates - Quantified improvements for each step
Difficulty indicators - Easy/Medium/Hard classifications
Executable commands - Concrete commands to run
Estimated effort - Time estimates in hours

Before (Legacy format):

ACTION: Add tests and refactor
WHY: High complexity with low coverage
STEPS: Write tests, reduce complexity, verify improvements

After (Concise format):

PRIMARY ACTION: Add 8 tests for untested branches
ESTIMATED EFFORT: 2.5 hours

STEPS:
1. Add 8 tests for 70% coverage gap [Easy]
   Impact: +8 tests, reduce risk
   Commands: cargo test parse_complex_input::
            # Write focused tests covering critical paths

2. Extract complex branches into focused functions [Medium]
   Impact: -15 complexity
   Commands: cargo clippy -- -W clippy::cognitive_complexity

3. Verify tests pass and coverage improved [Easy]
   Impact: Confirmed +70% coverage
   Commands: cargo test --all
            # Run coverage tool to verify improvement

The new format helps you:

Prioritize which step to do first (ordered by impact)
Estimate how long the work will take
Execute with specific commands to run
Verify improvements with measurable impact

HTML Dashboard Output

Debtmap can generate an interactive HTML dashboard for visualizing code quality metrics, complexity analysis, and technical debt:

# Generate HTML dashboard
debtmap analyze . --format html > debtmap-report.html

# With test coverage integration
debtmap analyze . --lcov target/coverage/lcov.info --format html > report.html

Dashboard Features:

Interactive Charts - Visual representations of issue distribution, root causes, and complexity patterns
Sortable Tables - Filter and sort complex functions by cyclomatic complexity, cognitive complexity, or nesting depth
Search Functionality - Quick search across all detected issues and functions
Metrics Overview - Summary cards showing critical/high/medium/low priority items
Complexity Analysis - Scatter plot showing relationship between cyclomatic and cognitive complexity
God Object Detection - Dedicated table for architectural debt items
Responsive Design - Works on desktop, tablet, and mobile devices

Opening the Dashboard:

Simply open the generated HTML file in any modern web browser:

# macOS
open debtmap-report.html

# Linux
xdg-open debtmap-report.html

# Windows
start debtmap-report.html

The dashboard is a single self-contained HTML file with no external dependencies (uses CDN for Chart.js and Tailwind CSS), making it easy to share with your team or include in CI/CD artifacts.

📖 See the Getting Started Guide for detailed installation, examples, and next steps.

GitHub Actions Integration

Automate debtmap analysis in your CI/CD pipeline with the debtmap GitHub Action.

Quick Setup

Add debtmap analysis to your workflow:

name: Code Quality

on: [push, pull_request]

jobs:
  debtmap:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: iepathos/debtmap-action@v1
        with:
          format: 'json'
          output: 'debtmap-report.json'

With Coverage Analysis

Combine with test coverage for comprehensive risk assessment:

name: Code Quality with Coverage

on: [push, pull_request]

jobs:
  analyze:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4

      - name: Generate coverage
        run: |
          cargo tarpaulin --out lcov --output-dir target/coverage

      - uses: iepathos/debtmap-action@v1
        with:
          coverage-file: 'target/coverage/lcov.info'
          format: 'json'
          output: 'debtmap-report.json'

Coverage Matching for Trait Methods (Rust)

Debtmap automatically handles coverage matching for Rust trait implementation methods, which can have different names between the AST analysis and LCOV coverage data.

How it works:

When analyzing Rust code, debtmap stores trait methods with their full qualified names (e.g., RecursiveMatchDetector::visit_expr), but LCOV often stores them with just the method name (e.g., visit_expr) after symbol demangling.

Debtmap tries multiple name variants automatically:

Full qualified name: RecursiveMatchDetector::visit_expr
Method name only: visit_expr
Trait-qualified name: Visit::visit_expr

Benefits:

✓ No false-positive "no coverage data" reports for trait methods
✓ Correctly reports coverage for syn::visit::Visit, std::fmt::Display, and other trait implementations
✓ Works automatically - no configuration needed
✓ Minimal performance impact (<2% overhead)

Example:

# Generate coverage
cargo llvm-cov --lcov --output-path target/coverage/lcov.info

# Analyze with coverage - trait methods automatically matched
debtmap analyze . --coverage-file target/coverage/lcov.info

# Verify specific trait method coverage
debtmap explain-coverage . \
  --coverage-file target/coverage/lcov.info \
  --function visit_expr \
  --file src/complexity/recursive_detector.rs

See Spec 181 for technical details.

Enforce Quality Gates

Fail builds when quality thresholds are exceeded:

- uses: iepathos/debtmap-action@v1
  with:
    max-complexity-density: '10.0'
    max-dependency-density: '5.0'
    min-test-density: '2.0'
    fail-on-violation: 'true'

📖 See the debtmap-action repository for complete documentation and configuration options.

Key Features

Coverage-Risk Correlation - Combines complexity with test coverage to prioritize genuinely risky code
Multi-Factor Analysis - Analyzes complexity, coverage, dependencies, and call graphs for comprehensive scoring
Reduced False Positives - Uses entropy analysis and pattern detection to distinguish genuine complexity from repetitive patterns (reduces false positives by up to 70%)
Test File Detection - Automatically identifies test files across languages and applies context-aware scoring adjustments
Actionable Recommendations - Specific guidance with quantified impact metrics
Multi-language Support - Full Rust support, partial Python/JavaScript/TypeScript
Fast Performance - 10-100x faster than Java/Python-based competitors (written in Rust with parallel processing)
Language-Agnostic Coverage - Works with any tool generating LCOV format
Context-Aware Analysis - Understands entry points, call graphs, and testing patterns
Free & Open Source - MIT licensed, no enterprise pricing required

📖 See the Getting Started Guide for complete feature documentation and examples.

Parallel Batch Analysis

Debtmap uses the traverse pattern from functional programming for efficient parallel file analysis. This approach provides:

Parallel processing - Analyze multiple files concurrently using all CPU cores
Error accumulation - Collect ALL errors instead of failing at the first one
Configurable batching - Control batch sizes for large codebases

Basic Usage

use debtmap::analyzers::batch::{analyze_files_effect, validate_files};
use debtmap::effects::{run_effect, run_validation};

// Parallel analysis with Effect pattern
let files = vec!["src/main.rs".into(), "src/lib.rs".into()];
let results = run_effect(
    analyze_files_effect(files),
    config,
)?;

// Process results
for result in results {
    println!("{}: {} functions analyzed",
        result.path.display(),
        result.metrics.complexity.functions.len()
    );
}

Validation with Error Accumulation

When you need comprehensive error reporting (showing ALL issues, not just the first):

use debtmap::analyzers::batch::validate_files;
use debtmap::effects::run_validation;

let paths = vec!["file1.rs".into(), "file2.rs".into(), "file3.rs".into()];
match run_validation(validate_files(&paths)) {
    Ok(validated_files) => {
        // All files are valid, proceed with analysis
        println!("Validated {} files", validated_files.len());
    }
    Err(all_errors) => {
        // Show ALL validation errors to user
        eprintln!("Found {} validation errors:", all_errors.len());
        for error in all_errors {
            eprintln!("  - {}", error);
        }
    }
}

Configuration

Control parallel processing behavior in .debtmap.toml:

[batch_analysis]
# Enable/disable parallel processing
parallelism.enabled = true

# Batch size for chunked processing (default: auto-detected based on CPU count)
parallelism.batch_size = 50

# Collect timing information for each file
collect_timing = true

Or via CLI:

# Analyze with timing information
debtmap analyze . --collect-timing

# Force sequential processing (useful for debugging)
debtmap analyze . --sequential

Performance Characteristics

File Count	Speedup (8 cores)	Memory
10 files	~2x	Minimal
50 files	~4x	Low
100+ files	~6-7x	Moderate

The traverse pattern batches files to balance parallelism with memory usage. For very large codebases (1000+ files), files are processed in configurable chunks.

When to Use Each Pattern

Pattern	Use Case
`analyze_files_effect`	Normal analysis - fail fast on errors
`validate_files`	Pre-validation - collect ALL errors for user feedback
`validate_and_analyze_files`	Combined - validate first, then analyze valid files

Advanced Features

God Object Detection

Debtmap identifies classes and modules with too many responsibilities using purity-weighted scoring that rewards functional programming patterns.

📖 Read more: God Object Detection

Understanding GOD OBJECT vs GOD MODULE

Debtmap distinguishes between two different organizational anti-patterns:

GOD OBJECT - A single struct/class with too many methods and fields:

Classification: >20 methods AND >5 fields on one struct/class
Problem: One class doing too much, methods share mutable state
Example output: GOD OBJECT: UserController (52 methods, 8 fields)
Fix: Extract responsibilities into focused classes

GOD MODULE - A file with too many diverse functions:

Classification: >20 module-level functions, but NOT a god object
Problem: Module lacks cohesion, contains unrelated utilities
Example output: GOD MODULE (47 module functions)
Fix: Split into cohesive submodules by domain

How to interpret the output:

When debtmap detects a god object, you'll see:

#3 SCORE: 7.5 [HIGH]
├─ GOD OBJECT: src/controller.rs
├─ TYPE: UserController (52 methods, 8 fields)
├─ ACTION: Extract responsibilities into focused classes
└─ WHY: Single class with too many methods and fields

The key indicators:

Methods: Number of methods on the dominant struct
Fields: Number of fields in that struct
This means refactor the specific struct, not the whole file

When debtmap detects a god module, you'll see:

#5 SCORE: 6.8 [HIGH]
├─ GOD MODULE: src/utils.rs
├─ TYPE: Module with 47 diverse functions
├─ ACTION: Split into cohesive submodules by domain
└─ WHY: Module lacks focus, contains unrelated utilities

The key indicators:

Module Functions: Total count of module-level functions
This means reorganize the file's functions into multiple focused modules

Quick Decision Guide:

See "GOD OBJECT"? Extract that specific class into smaller classes
See "GOD MODULE"? Split the file's functions into multiple focused modules
Both can appear in the same codebase for different files

Smart Refactoring Recommendations

Debtmap provides tailored recommendations based on your file's characteristics:

Struct-Heavy Modules (many type definitions):

Detection criteria: 5+ structs with 3+ semantic domains, struct-to-function ratio > 0.3
Recommendation style: Domain-based organization
Example: A config.rs file with ScoreConfig, ThresholdConfig, DetectionConfig will be recommended to split into:
- config/scoring.rs - Score-related structures
- config/thresholds.rs - Threshold-related structures
- config/detection.rs - Detection-related structures
Why: Groups related types together for better semantic cohesion

Method-Heavy Modules (many functions):

Detection criteria: Does not meet struct-heavy criteria
Recommendation style: Responsibility-based organization
Example: A utility file with diverse functions will be recommended to split by responsibility:
- parsing.rs - Input parsing functions
- formatting.rs - Output formatting functions
- validation.rs - Validation functions
Why: Separates different functional concerns for clarity

Severity Levels:

Critical: God object with cross-domain mixing (immediate action recommended)
High: Significant complexity or size issues (priority refactoring)
Medium: Proactive improvement opportunity (approaching thresholds)
Low: Informational suggestions (minor improvements)

Domain Diversity Analysis

For struct-heavy modules, debtmap performs domain diversity analysis to identify cross-domain mixing patterns that violate the single responsibility principle.

How It Works:

Analyzes struct naming patterns to identify semantic domains (e.g., "Config", "Error", "Handler")
Calculates domain diversity scores based on struct distribution across domains
Assigns severity levels from OK to CRITICAL based on diversity

Severity Levels:

OK: Single domain or closely related domains (diversity ≤ 0.4)
MODERATE: Some domain mixing (0.4 < diversity ≤ 0.6)
HIGH: Significant cross-domain concerns (0.6 < diversity ≤ 0.75)
CRITICAL: Severe domain mixing (diversity > 0.75)

Example Output:

WHY THIS MATTERS: This module contains 12 structs across 4 distinct domains.
Cross-domain mixing (Severity: CRITICAL) violates single responsibility
principle and increases maintenance complexity.

DOMAIN DIVERSITY ANALYSIS (Spec 140):
Severity: CRITICAL - 12 structs across 4 domains

Domain Distribution:
  - Configuration: 5 structs (42%)
    Examples: AppConfig, DatabaseConfig, CacheConfig
  - Error Handling: 4 structs (33%)
    Examples: ParseError, ValidationError, NetworkError
  - Request Processing: 2 structs (17%)
    Examples: HttpRequest, ApiResponse
  - Caching: 1 structs (8%)
    Examples: CacheEntry

Recommendation: Split into domain-focused modules for better cohesion

This analysis helps you understand exactly why a module should be split and provides clear guidance on how to organize the extracted modules by domain.

Example recommendation output:

GOD OBJECT DETECTED: src/config.rs (10 structs across 3 domains)
  Recommendation: Split by semantic domain
  Severity: High

  Suggested splits:
    1. config/scoring.rs
       Structs: ScoreConfig, ScoreCalculator, ScoreValidator
       Estimated lines: ~150

    2. config/thresholds.rs
       Structs: ThresholdConfig, ThresholdValidator, ThresholdManager, ThresholdFactory
       Estimated lines: ~200

Semantic Module Naming

When splitting god objects, debtmap uses intelligent semantic naming to generate descriptive, meaningful module names based on the methods in each split. This eliminates generic names like utils, misc, or helpers and ensures each split has a clear, specific identity.

How It Works:

Domain Term Extraction: Analyzes method names to find common domain terms (e.g., "coverage", "metrics", "config")
Behavioral Pattern Recognition: Identifies behavioral patterns like "formatting", "validation", "parsing", "computation"
Specificity Scoring: Ensures names are descriptive, rejecting generic terms
Uniqueness Validation: Guarantees no filename collisions across splits

Naming Strategies:

Domain Terms: Extracts dominant terms from method names
- Methods: format_coverage_status, format_coverage_factor, calculate_coverage_percentage
- Generated name: coverage (confidence: 0.85)
Behavioral Patterns: Recognizes common software patterns
- Methods: validate_index, validate_data, validate_config
- Generated name: validation (confidence: 0.75)
Descriptive Fallback: When no clear pattern emerges, generates meaningful placeholders
- Methods: do_something, handle_stuff
- Generated name: needs_review_group_1 (confidence: 0.4)

Confidence Scoring:

High (0.7-1.0): Clear, unambiguous pattern detected
Medium (0.5-0.7): Reasonable pattern with some uncertainty
Low (0.4-0.5): Fallback name, manual review recommended
Rejected (<0.4): Name too generic, alternative generated

Example Output:

GOD OBJECT DETECTED: src/data_manager.rs (24 methods)

  Suggested splits:
    1. data_manager/formatting.rs (confidence: 0.85)
       Methods: format_output, format_summary, format_report
       Responsibility: Output formatting operations

    2. data_manager/validation.rs (confidence: 0.78)
       Methods: validate_index, validate_data, validate_config
       Responsibility: Input validation

    3. data_manager/parsing.rs (confidence: 0.72)
       Methods: parse_input, parse_config, parse_json
       Responsibility: Data parsing operations

Alternative Names: Each split includes up to 3 name candidates ranked by confidence, allowing you to choose the most appropriate name for your codebase conventions.

Framework Pattern Detection

Debtmap identifies framework-specific code patterns across Rust, Python, JavaScript, and TypeScript, improving the accuracy of responsibility classification and helping distinguish framework boilerplate from application logic.

Supported Frameworks:

Rust: Axum, Actix-Web, Tokio, Diesel, Clap
Python: FastAPI, Flask, Django, Pytest, SQLAlchemy, Click, Celery
JavaScript/TypeScript: Express.js, Fastify, React, Jest, Mocha, NestJS, Prisma

How It Works:

Framework patterns are detected using a combination of:

Import/require statements
Decorators and attributes
Function signatures and parameters
Return types and naming conventions
File path patterns

Example Detection:

// Axum Web Handler - Detected as "HTTP Request Handler"
async fn get_user(Path(user_id): Path<u32>) -> Json<User> {
    // ...
}

# FastAPI Route - Detected as "HTTP Request Handler"
@app.get("/users/{user_id}")
async def get_user(user_id: int) -> User:
    # ...

// React Component - Detected as "UI Component"
function UserProfile({ userId }) {
    return <div>Profile for {userId}</div>;
}

Custom Pattern Configuration:

You can add custom framework patterns by creating a framework_patterns.toml file in your project root:

[rust.web.your_framework]
name = "Your Framework"
category = "HTTP Request Handler"
patterns = [
    { type = "import", pattern = "your_framework::" },
    { type = "parameter", pattern = "Request<" },
    { type = "return_type", pattern = "Response" },
]

Pattern types available:

import - Match import/use statements
decorator - Match Python/TypeScript decorators
attribute - Match Rust attributes (#[...])
derive - Match Rust derive macros
parameter - Match function parameter types
return_type - Match function return types
name - Match function names (regex supported)
call - Match function calls in body
file_path - Match file paths (regex supported)

Benefits:

Better Responsibility Classification: Framework handlers are correctly categorized instead of being flagged as generic "I/O" operations
Reduced False Positives: Test functions and framework boilerplate are properly identified
Context-Aware Analysis: Understanding framework patterns helps debtmap provide more accurate complexity assessments

Test File Detection and Context-Aware Scoring

Debtmap automatically identifies test files and test functions across multiple languages, then applies context-aware scoring adjustments to reduce false positives from test-specific patterns.

Multi-Language Test Detection:

Debtmap detects test files using language-specific patterns:

Rust: #[test], #[cfg(test)], files in tests/ directory, _test.rs suffix
Python: test_*.py, *_test.py, unittest, pytest imports, def test_*() functions
JavaScript/TypeScript: *.test.js, *.spec.ts, Jest/Mocha imports, describe()/it() blocks
General: Files in tests/, test/, __tests__/ directories

Context-Aware Scoring:

When debtmap identifies a test file or test function, it automatically:

Reduces complexity penalties - Test code often has high cyclomatic complexity (many branches for edge cases) but is maintainable
Adjusts priority levels - Test debt is scored lower priority than production code debt
Changes coverage expectations - Test files don't need test coverage themselves
Provides test-specific recommendations - Suggests test refactoring patterns instead of production refactoring patterns

Example Output:

#7 SCORE: 4.2 [MEDIUM]
├─ TEST CODE: ./tests/integration_test.rs:125 test_complex_workflow()
├─ COMPLEXITY: cyclomatic=12, cognitive=8 (test-adjusted)
├─ ACTION: Extract test helper functions for reusability
└─ WHY: Test complexity is acceptable but helpers improve maintainability

Benefits:

Fewer false positives - Test code complexity doesn't dominate production priorities
Better recommendations - Test-specific refactoring guidance
Language consistency - Works across Rust, Python, JavaScript, and TypeScript
Automatic detection - No configuration needed for standard test patterns

📖 Read more: Testing Guide

Pattern Detection

Automatically detects common design patterns (Observer, Factory, Singleton, Strategy, etc.) with configurable confidence thresholds.

📖 Read more: Analysis Guide

Pure Mapping Pattern Detection

Reduces false positives from exhaustive match expressions that are actually simple and maintainable. Debtmap recognizes pure mapping patterns - match statements that transform input to output without side effects - and adjusts complexity scores accordingly.

What's a pure mapping pattern?

fn status_to_string(status: Status) -> &'static str {
    match status {
        Status::Success => "success",
        Status::Pending => "pending",
        Status::Failed => "failed",
        Status::Cancelled => "cancelled",
        // ... many more cases
    }
}

This function has high cyclomatic complexity (one branch per case), but it's simple to maintain because:

Each branch is independent and straightforward
No mutation or side effects occur
The pattern is predictable and easy to understand
Adding new cases requires minimal changes

Impact: By recognizing these patterns, debtmap reduces complexity scores by up to 30% for pure mapping functions, preventing them from incorrectly appearing as high-priority refactoring targets.

Configuration: Customize detection thresholds in .debtmap.toml:

[mapping_patterns]
enabled = true                      # Enable mapping pattern detection
complexity_reduction = 0.30         # Reduce complexity by 30%
min_branches = 3                    # Minimum match arms to consider

📖 Read more: Configuration Guide

Role-Based Coverage Expectations

Debtmap recognizes that different types of functions have different testing priorities. Instead of applying a uniform 80% coverage target to all code, it uses role-specific expectations that reflect real-world testing best practices.

Default Coverage Expectations by Role:

Function Role	Target	Why
Pure Logic	90-100%	Easy to test, high ROI
Business Logic	80-95%	Critical functionality
Validation	85-98%	Must be correct
State Management	75-90%	Complex behavior
Error Handling	70-90%	Important paths
I/O Operations	60-80%	Often integration tested
Configuration	60-80%	Lower risk
Orchestration	65-85%	Coordinating functions
Utilities	75-95%	Should be reliable
Initialization	50-75%	Lower priority
Performance	40-60%	Optimization code
Debug/Development	20-40%	Development-only code

How it works:

When debtmap identifies a function with low coverage, it considers the function's role:

A pure function with 70% coverage gets flagged (below 90% target)
A debug function with 70% coverage is fine (above 30% target)

Example output:

#2 SCORE: 7.2 [HIGH]
├─ TEST GAP: ./src/calc.rs:42 compute_price()
├─ COVERAGE: 65% (expected: 90% for Pure functions) 🟠
├─ ACTION: Add 8 unit tests to reach target
└─ WHY: Pure logic is easy to test and high-value

Customize expectations in .debtmap.toml:

[coverage_expectations]
pure = { min = 90.0, target = 95.0, max = 100.0 }
business_logic = { min = 80.0, target = 90.0, max = 95.0 }
debug = { min = 20.0, target = 30.0, max = 40.0 }

Manual role override:

You can override automatic role detection using doc comments:

/// Calculate user discount
/// @debtmap-role: BusinessLogic
fn calculate_discount(user: &User) -> f64 {
    // debtmap will use BusinessLogic expectations (80-95%)
}

Coverage gap severity indicators:

🟢 Meets or exceeds target
🟡 Between min and target (minor gap)
🟠 Below min but above 50% of min (moderate gap)
🔴 Critically low (below 50% of min)

📖 Read more: Coverage Integration Guide

Complexity Scoring

Debtmap uses weighted complexity scoring that combines cyclomatic and cognitive complexity metrics with configurable weights. This approach provides more accurate prioritization by emphasizing cognitive complexity, which research shows correlates better with bug density and maintenance difficulty.

Why cognitive complexity matters:

Cyclomatic complexity counts control flow branches (if, while, for, etc.)
Cognitive complexity measures how hard code is to understand (nested conditions, breaks in linear flow)
A function can have high cyclomatic but low cognitive complexity (e.g., a simple switch statement with many cases)
Conversely, deeply nested conditionals have high cognitive complexity even with few branches

Default weights:

70% cognitive complexity - Emphasizes human understanding difficulty
30% cyclomatic complexity - Still considers control flow complexity
Weights must sum to 1.0 and can be customized per project

Weighted score calculation:

Normalize both metrics to 0-100 scale (default: cyclomatic max=50, cognitive max=100)
Apply weights: score = (0.3 × normalized_cyclomatic) + (0.7 × normalized_cognitive)
Display as: cyclomatic=15, cognitive=3 → weighted=11.1 (cognitive-driven)

Configuration in .debtmap.toml:

[complexity_weights]
# Customize weights (must sum to 1.0)
cyclomatic = 0.3
cognitive = 0.7

# Adjust normalization based on your codebase
max_cyclomatic = 50.0
max_cognitive = 100.0

Benefits:

Reduces false positives from simple repetitive patterns (e.g., mapping functions)
Prioritizes deeply nested logic that's truly hard to understand
Transparent scoring shows all metrics and the dominant driver
Configurable for different project needs

📖 Read more: Analysis Guide

Suppression Patterns

Flexible suppression via inline comments or configuration files.

📖 Read more: Suppression Patterns

Contributing

We welcome contributions! This is an early-stage project, so there's plenty of room for improvement.

📖 See the Contributing Guide for detailed development setup and contribution guidelines.

Please note that this project is released with a Code of Conduct. By participating in this project you agree to abide by its terms.

Areas for Contribution

Language support - Add analyzers for Go, Java, etc.
New metrics - Implement additional complexity or quality metrics
Speed - Optimize analysis algorithms
Documentation - Improve docs and add examples
Testing - Expand test coverage

Development

This project uses Just for task automation.

# Common development tasks
just test        # Run all tests
just fmt         # Format code
just lint        # Run clippy linter
just check       # Quick syntax check
just dev         # Run in development mode
just watch       # Run with hot reloading

# CI and quality checks
just ci          # Run all CI checks locally
just coverage    # Generate test coverage report (uses cargo-llvm-cov)

# See all available commands
just --list

Automated Technical Debt Reduction

📖 See the Prodigy Integration Guide for detailed information on using Prodigy and Claude Code for automated debt reduction.

We use prodigy for automated technical debt reduction through AI-driven workflows:

# Run automated debt reduction (5 iterations)
prodigy run workflows/debtmap.yml -yn 5

This command creates an isolated git worktree, runs iterations of automated improvements, validates changes, and commits with detailed metrics.

License

MIT License - see LICENSE file for details

Dependency Licensing Note

Debtmap includes Python parsing functionality via rustpython-parser, which depends on malachite (LGPL-3.0 licensed) for arbitrary-precision arithmetic. This LGPL dependency is used only for Python AST parsing and does not affect the MIT licensing of debtmap itself. For use cases requiring strict MIT-only dependencies, Python support can be disabled or replaced with an alternative parser.

Debugging Call Graph Issues

Debtmap includes powerful debugging and diagnostic tools for troubleshooting call graph analysis and understanding function relationship detection.

Debug Call Graph Resolution

View detailed information about how functions are resolved and linked in the call graph:

# Enable debug mode for call graph analysis
debtmap analyze . --debug-call-graph

# Output debug information in JSON format
debtmap analyze . --debug-call-graph --debug-format json

# Trace specific functions to see their resolution details
debtmap analyze . --debug-call-graph --trace-function my_function --trace-function other_function

Debug output includes:

Resolution statistics (success rate, failure reasons)
Strategy performance (exact match, fuzzy matching, etc.)
Timing percentiles (p50, p95, p99) for performance analysis
Failed resolutions with detailed candidate information
Recommendations for improving resolution accuracy

Validate Call Graph Structure

Check the structural integrity and health of the generated call graph:

# Run validation checks on call graph
debtmap analyze . --validate-call-graph

# Combine validation with debug output
debtmap analyze . --validate-call-graph --debug-call-graph

Validation checks:

Structural Issues: Detects dangling edges, orphaned nodes, and duplicate functions
Heuristic Warnings: Identifies suspicious patterns like unusually high fan-in/fan-out
Health Score: Overall graph quality score (0-100) based on detected issues
Detailed Reports: Shows specific issues with file locations and function names

View Call Graph Statistics

Get quick statistics about call graph size and structure:

# Show call graph statistics only (fast, minimal output)
debtmap analyze . --call-graph-stats-only

Statistics include:

Total number of functions analyzed
Total number of function calls detected
Average calls per function (graph density)

Common Use Cases

Debugging unresolved function calls:

# See why specific functions aren't being linked
debtmap analyze . --debug-call-graph --trace-function problematic_function

Validating analysis quality:

# Check for structural problems in call graph
debtmap analyze . --validate-call-graph

Performance profiling:

# See timing breakdown of call resolution
debtmap analyze . --debug-call-graph --debug-format json

Combining with normal analysis:

# Run full analysis with debugging enabled
debtmap analyze . --lcov coverage.info --debug-call-graph --validate-call-graph

Interpreting Debug Output

Health Score:

95-100: Excellent - Very few unresolved calls
85-94: Good - Acceptable resolution rate
<85: Needs attention - High number of unresolved calls

Resolution Strategies:

Exact: Exact function name match (highest confidence)
Fuzzy: Qualified name match (e.g., Module::function)
NameOnly: Base name match (lowest confidence, may have ambiguity)

Common Issues:

Dangling Edges: References to non-existent functions (potential parser bugs)
Orphaned Nodes: Functions with no connections (may indicate missed calls)
High Fan-Out: Functions calling many others (potential god objects)
High Fan-In: Functions called by many others (potential bottlenecks)

Performance Considerations

Debug and validation modes add minimal overhead (<20% typically) and can be used in CI/CD pipelines. For large codebases (>1000 files), consider:

Using --call-graph-stats-only for quick health checks
Limiting --trace-function to specific problem areas
Running full debug analysis periodically rather than on every build

Viewing Dependency Information

Debtmap displays caller/callee relationships for each technical debt item, helping you understand the impact and reach of functions that need attention.

Dependency Display in Output

When running analysis with default verbosity (-v), each debt item includes a DEPENDENCIES section showing:

#1 SCORE: 8.9 [CRITICAL]
├─ TEST GAP: ./src/parser.rs:38 parse_complex_input()
├─ ACTION: Add 6 unit tests for full coverage
├─ IMPACT: -3.7 risk reduction
├─ DEPENDENCIES:
|  |- Called by (3):
|       ⬆ validate_input
|       ⬆ process_request
|       ⬆ handle_api_call
|  |- Calls (2):
|       ⬇ tokenize
|       ⬇ validate_syntax
└─ WHY: Complex logic (cyclomatic=6) with 0% test coverage

What the dependency information shows:

Called by (callers): Functions that depend on this function (upward arrow ⬆)
Calls (callees): Functions this function depends on (downward arrow ⬇)
Counts are shown in parentheses (e.g., "(3)" means 3 callers)

Configuring Dependency Display

Control how many dependencies are shown using CLI flags:

# Limit callers and callees displayed (default: 5 each)
debtmap analyze . --max-callers 10 --max-callees 10

# Show external crate calls (hidden by default)
debtmap analyze . --show-external-calls

# Show standard library calls (hidden by default)
debtmap analyze . --show-std-lib-calls

# Hide all dependency information
debtmap analyze . --no-dependencies

Configuration File

Add dependency display settings to .debtmap.toml:

[output.dependencies]
max_callers = 10        # Maximum callers to display (default: 5)
max_callees = 10        # Maximum callees to display (default: 5)
show_external = false   # Show external crate calls (default: false)
show_std_lib = false    # Show stdlib calls (default: false)

Understanding Dependency Impact

Dependency information helps prioritize refactoring:

High caller count → Changes affect many parts of codebase (higher refactoring risk)
High callee count → Function has many dependencies (higher complexity)
Entry points (few/no callers) → Good starting points for testing
Leaf functions (few/no callees) → Easier to test in isolation

CI/CD Integration with Density-Based Validation

Debtmap supports density-based validation metrics that work consistently across projects of any size. Unlike traditional absolute thresholds (e.g., "max complexity of 1000"), density metrics normalize by codebase size, making them ideal for CI/CD automation.

Why Density-Based Metrics?

Traditional metrics fail across different project sizes:

A 1,000-line project with complexity 500 → 50% of threshold
A 100,000-line project with complexity 5,000 → 500% of threshold

Density metrics solve this by measuring per-line or per-function rates:

Complexity density = total_complexity / total_functions
Same threshold works for any project size
Quality standards remain consistent as code grows

Available Density Metrics

Metric	Formula	Good Threshold	Description
Complexity Density	`total_complexity / total_functions`	< 10.0	Average complexity per function
Dependency Density	`(dependencies / lines) * 1000`	< 5.0	Dependencies per 1,000 lines
Test Density	`(tests / lines) * 100`	> 2.0	Tests per 100 lines

Quick Start: GitHub Actions

Add density-based validation to your CI pipeline:

name: Code Quality

on: [push, pull_request]

jobs:
  quality:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4

      - name: Install debtmap
        run: curl -sSL https://raw.githubusercontent.com/iepathos/debtmap/master/install.sh | bash

      - name: Validate code quality
        run: |
          debtmap analyze . \
            --max-complexity-density 10.0 \
            --max-dependency-density 5.0 \
            --min-test-density 2.0

Benefits:

No threshold adjustments needed as your codebase grows
Catches quality degradation early
Consistent standards across all projects
Predictable CI/CD behavior

Setting Appropriate Thresholds

For New Projects

Start with industry best practices:

debtmap analyze . \
  --max-complexity-density 8.0 \    # Excellent: simple functions
  --max-dependency-density 3.0 \    # Minimal dependencies
  --min-test-density 2.5            # Comprehensive tests

For Existing Projects

Baseline analysis - Understand current state:

debtmap analyze . --density-metrics > baseline.json

Set initial thresholds - Current values + 20% buffer:

# Example: Current complexity density is 12.5
debtmap analyze . --max-complexity-density 15.0

Gradual improvement - Tighten thresholds quarterly:

# Q1: Stabilize
--max-complexity-density 15.0

# Q2: Improve
--max-complexity-density 13.0

# Q3: Approach best practices
--max-complexity-density 10.0

# Q4: Maintain excellence
--max-complexity-density 8.0

CI/CD Configuration Examples

GitHub Actions - Pull Request Validation

name: PR Quality Check

on: pull_request

jobs:
  quality:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
        with:
          fetch-depth: 0  # Full history for delta comparison

      - name: Install debtmap
        run: curl -sSL https://raw.githubusercontent.com/iepathos/debtmap/master/install.sh | bash

      - name: Analyze base branch
        run: |
          git checkout ${{ github.base_ref }}
          debtmap analyze . --density-metrics --format json > base.json

      - name: Analyze PR branch
        run: |
          git checkout ${{ github.head_ref }}
          debtmap analyze . --density-metrics --format json > pr.json

      - name: Check density delta
        run: |
          BASE_DENSITY=$(jq '.density_metrics.complexity_density' base.json)
          PR_DENSITY=$(jq '.density_metrics.complexity_density' pr.json)
          DELTA=$(echo "$PR_DENSITY - $BASE_DENSITY" | bc)

          if (( $(echo "$DELTA > 0.5" | bc -l) )); then
            echo "❌ Complexity density increased by $DELTA"
            exit 1
          fi

          echo "✅ Complexity density change: $DELTA"

      - name: Enforce absolute limits
        run: |
          debtmap analyze . \
            --max-complexity-density 10.0 \
            --max-dependency-density 5.0 \
            --min-test-density 2.0

GitLab CI - Multi-Stage Validation

stages:
  - analyze
  - validate

code_analysis:
  stage: analyze
  script:
    - curl -sSL https://raw.githubusercontent.com/iepathos/debtmap/master/install.sh | bash
    - debtmap analyze . --density-metrics --format json > metrics.json
  artifacts:
    paths:
      - metrics.json
    expire_in: 1 week

quality_gates:
  stage: validate
  dependencies:
    - code_analysis
  script:
    - debtmap analyze . --max-complexity-density 10.0 --max-dependency-density 5.0 --min-test-density 2.0
  only:
    - merge_requests
    - master

CircleCI - Density Tracking

version: 2.1

jobs:
  quality_check:
    docker:
      - image: cimg/rust:1.75
    steps:
      - checkout
      - run:
          name: Install debtmap
          command: curl -sSL https://raw.githubusercontent.com/iepathos/debtmap/master/install.sh | bash

      - run:
          name: Analyze and validate
          command: |
            debtmap analyze . \
              --density-metrics \
              --max-complexity-density 10.0 \
              --max-dependency-density 5.0 \
              --min-test-density 2.0 \
              --format json > /tmp/metrics.json

      - store_artifacts:
          path: /tmp/metrics.json
          destination: code-metrics

workflows:
  version: 2
  build:
    jobs:
      - quality_check

Advanced CI/CD Patterns

Progressive Tightening

Automatically adjust thresholds based on historical data:

#!/bin/bash
# progressive-quality.sh

CURRENT_DENSITY=$(debtmap analyze . --density-metrics --format json | jq '.density_metrics.complexity_density')
HISTORICAL_AVG=12.5  # From last 30 days

if (( $(echo "$CURRENT_DENSITY < $HISTORICAL_AVG" | bc -l) )); then
  # Quality improved - tighten threshold
  NEW_THRESHOLD=$(echo "$CURRENT_DENSITY * 1.1" | bc)
  echo "✅ Quality improved! New threshold: $NEW_THRESHOLD"
else
  # Use current average
  NEW_THRESHOLD=$HISTORICAL_AVG
fi

debtmap analyze . --max-complexity-density "$NEW_THRESHOLD"

Multi-Environment Thresholds

Different standards for different branches:

- name: Validate code quality
  run: |
    if [ "${{ github.ref }}" == "refs/heads/main" ]; then
      # Strict for production
      debtmap analyze . --max-complexity-density 8.0
    elif [ "${{ github.ref }}" == "refs/heads/develop" ]; then
      # Moderate for development
      debtmap analyze . --max-complexity-density 10.0
    else
      # Lenient for feature branches
      debtmap analyze . --max-complexity-density 12.0
    fi

Team-Specific Thresholds

Different teams, different standards:

- name: Validate code quality
  run: |
    # Detect which team owns the changed files
    TEAM=$(git diff --name-only ${{ github.base_ref }} | xargs dirname | sort -u | head -1)

    case "$TEAM" in
      "src/core")
        # Core team: strict standards
        debtmap analyze src/core --max-complexity-density 6.0
        ;;
      "src/features")
        # Feature teams: moderate standards
        debtmap analyze src/features --max-complexity-density 10.0
        ;;
      *)
        # Default standards
        debtmap analyze . --max-complexity-density 8.0
        ;;
    esac

Monitoring Density Trends

Track density metrics over time to identify trends:

# Store metrics with timestamp
DATE=$(date +%Y-%m-%d)
debtmap analyze . --density-metrics --format json > "metrics-$DATE.json"

# Plot trend (requires jq and gnuplot)
for file in metrics-*.json; do
  DATE=$(echo "$file" | sed 's/metrics-\(.*\)\.json/\1/')
  DENSITY=$(jq '.density_metrics.complexity_density' "$file")
  echo "$DATE $DENSITY"
done | gnuplot -e "
  set terminal png;
  set output 'density-trend.png';
  plot '-' using 1:2 with lines title 'Complexity Density'
"

Troubleshooting CI/CD Integration

Issue: Thresholds fail on small codebases

Cause: Small projects have high variance in density metrics Solution: Require minimum codebase size:

LINES=$(find . -name "*.rs" | xargs wc -l | tail -1 | awk '{print $1}')
if [ "$LINES" -gt 1000 ]; then
  debtmap analyze . --max-complexity-density 10.0
else
  echo "⚠️  Codebase too small for density validation (${LINES} lines)"
fi

Issue: Density metrics fluctuate wildly

Cause: Including/excluding test files inconsistently Solution: Always exclude test files from production metrics:

debtmap analyze . \
  --exclude "**/tests/**" \
  --exclude "**/*_test.rs" \
  --max-complexity-density 10.0

Issue: Legacy code dominates metrics

Cause: Old code with high complexity affects overall density Solution: Analyze new and legacy code separately:

# Strict for new code
debtmap analyze src/new_features --max-complexity-density 8.0

# Lenient for legacy
debtmap analyze src/legacy --max-complexity-density 15.0

Migration Guide

For detailed information on migrating from scale-dependent to density-based validation, see the Validation Migration Guide.

The guide includes:

Why migrate and key benefits
Step-by-step migration process
Threshold selection guidelines
Example configurations for different project sizes
Common migration questions and troubleshooting

Benefits of Density-Based Metrics in Automation

✅ Size-independent: Same thresholds work for 1K or 1M lines ✅ Predictable: No surprise CI failures as code grows ✅ Meaningful: Measures actual code quality, not just size ✅ Actionable: Clear signals for refactoring priorities ✅ Maintainable: Set once, rarely need adjustment

Multi-Signal Responsibility Classification

Debtmap uses multi-signal aggregation to accurately classify function responsibilities, achieving ~88% accuracy compared to ~50% with name-based classification alone.

Signals

The classification system combines multiple independent signals:

Signal	Weight	Purpose
I/O Detection	35%	Identifies file, network, and database operations
Call Graph Analysis	25%	Detects orchestration and coordination patterns
Type Signatures	15%	Infers responsibility from parameter and return types
Name Heuristics	15%	Uses function naming conventions
Purity Analysis	5%	Identifies pure computation functions
Framework Patterns	5%	Detects framework-specific patterns (web handlers, tests, CLI)

Classification Categories

The system classifies functions into these responsibility categories:

I/O Operations:

File I/O, Network I/O, Database I/O, Configuration I/O

Handlers:

HTTP Request Handler, WebSocket Handler, CLI Handler, Database Handler

Computation:

Pure Computation, Validation, Transformation, Parsing, Formatting

Coordination:

Orchestration, Coordination, Error Handling

Testing:

Test Function

Accuracy & Validation

Baseline (name-only): ~50% accuracy
Multi-signal: ~88% accuracy (+38% improvement)
Validated against: 15+ manually classified test cases across all categories
Configuration: Tunable weights in aggregation_config.toml

Explainability

Each classification includes:

Primary category with confidence score
Evidence from each signal that contributed
Alternative classifications with their scores
Clear reasoning for troubleshooting misclassifications

Example output:

{
  "primary": "FileIO",
  "confidence": 0.82,
  "evidence": [
    {"signal": "io_detection", "contribution": 0.35, "description": "2 file ops"},
    {"signal": "name_heuristics", "contribution": 0.11, "description": "Name pattern: read_config"}
  ],
  "alternatives": [
    {"category": "ConfigurationIO", "score": 0.24}
  ]
}

Benefits

✅ Higher accuracy: 88% vs 50% name-based alone ✅ Reduced false positives: Multiple signals must agree ✅ Language-agnostic: Works across Rust, Python, JavaScript, TypeScript ✅ Explainable: Clear evidence trail for each classification ✅ Configurable: Adjust weights for your codebase's patterns ✅ Performance: <3% overhead with parallel processing

Roadmap

Language Support

Rust - Full support with AST parsing and macro expansion
Python - Full support via rustpython-parser
JavaScript/TypeScript - Full support via tree-sitter
Go - Planned
C/C++ - Planned
C# - Planned
Java - Planned

Core Features

Inline suppression comments
LCOV coverage integration with risk analysis
Risk-based testing prioritization
Comprehensive debt detection (20+ pattern types)
Security vulnerability detection
Resource management analysis
Code organization assessment
Testing quality evaluation
Historical trend tracking

Integrations

Acknowledgments

Built with excellent Rust crates including:

syn for Rust AST parsing
rustpython-parser for Python parsing
tree-sitter for JavaScript/TypeScript parsing
rayon for parallel processing
clap for CLI parsing

Note: This is a prototype tool under active development. Please report issues and feedback on GitHub. For detailed documentation, visit iepathos.github.io/debtmap.

Name		Name	Last commit message	Last commit date
Latest commit History 3,507 Commits
.claude		.claude
.github		.github
.prodigy		.prodigy
benches		benches
book		book
docs		docs
examples		examples
proptest-regressions		proptest-regressions
specs		specs
src		src
tests		tests
workflows		workflows
.coverage		.coverage
.debtmap-improved.toml		.debtmap-improved.toml
.debtmap.toml		.debtmap.toml
.debtmap.toml.example		.debtmap.toml.example
.gitignore		.gitignore
ARCHITECTURE.md		ARCHITECTURE.md
CLAUDE.md		CLAUDE.md
CODE_OF_CONDUCT.md		CODE_OF_CONDUCT.md
CONTRIBUTING.md		CONTRIBUTING.md
Cargo.lock		Cargo.lock
Cargo.toml		Cargo.toml
Justfile		Justfile
LICENSE		LICENSE
README.md		README.md
aggregation_config.toml		aggregation_config.toml
deny.toml		deny.toml
framework_patterns.toml		framework_patterns.toml
generate_test_corpus.sh		generate_test_corpus.sh
install.sh		install.sh
libtest_purity.rlib		libtest_purity.rlib
rust-toolchain.toml		rust-toolchain.toml

License

iepathos/debtmap

Folders and files

Latest commit

History

Repository files navigation

debtmap

Why Debtmap?

Documentation

Quick Links

Quick Start (3 Minutes)

Install

Analyze

Review Results

Concise Actionable Recommendations

HTML Dashboard Output

GitHub Actions Integration

Quick Setup

With Coverage Analysis

Coverage Matching for Trait Methods (Rust)

Enforce Quality Gates

Key Features

Parallel Batch Analysis

Basic Usage

Validation with Error Accumulation

Configuration

Performance Characteristics

When to Use Each Pattern

Advanced Features

God Object Detection

Understanding GOD OBJECT vs GOD MODULE

Smart Refactoring Recommendations

Domain Diversity Analysis

Semantic Module Naming

Framework Pattern Detection

Test File Detection and Context-Aware Scoring

Pattern Detection

Pure Mapping Pattern Detection

Role-Based Coverage Expectations

Complexity Scoring

Suppression Patterns

Contributing

Areas for Contribution

Development

Automated Technical Debt Reduction

License

Dependency Licensing Note

Debugging Call Graph Issues

Debug Call Graph Resolution

Validate Call Graph Structure

View Call Graph Statistics

Common Use Cases

Interpreting Debug Output

Performance Considerations

Viewing Dependency Information

Dependency Display in Output

Configuring Dependency Display

Configuration File

Understanding Dependency Impact

CI/CD Integration with Density-Based Validation

Why Density-Based Metrics?

Available Density Metrics

Quick Start: GitHub Actions

Setting Appropriate Thresholds

For New Projects

For Existing Projects

CI/CD Configuration Examples

GitHub Actions - Pull Request Validation

GitLab CI - Multi-Stage Validation

CircleCI - Density Tracking

Advanced CI/CD Patterns

Progressive Tightening

Multi-Environment Thresholds

Team-Specific Thresholds

Monitoring Density Trends

Troubleshooting CI/CD Integration

Issue: Thresholds fail on small codebases

Issue: Density metrics fluctuate wildly

Issue: Legacy code dominates metrics

Migration Guide

Packages