🏗 Architecture

VS Code Extension (TypeScript)
        ↓
Python Analysis Engine
        ↓
8-Stage Autonomous Pipeline
        ↓
LLM Integration Layer
        ↓
Generated Tests + Diagnostics + Reports

RepoRelic combines a TypeScript VS Code extension with a Python-powered analysis engine to perform deep autonomous repository analysis.

⚡ Analysis Pipeline

Understand Codebase
        ↓
Static Analysis
        ↓
Dependency Graph
        ↓
Knowledge Graph
        ↓
Test Generation
        ↓
Execution Engine
        ↓
Failure Diagnosis
        ↓
Output Report

Each stage enriches repository understanding and feeds structured intelligence into later stages.

📁 Generated Output Structure

After analysis, RepoRelic creates:

your_project/
├── .reporelic/
│   ├── generated_tests/
│   │   ├── test_example.py
│   │   └── ...
│   │
│   ├── knowledge_graph.json
│   ├── report.md
│   └── diagnostics/
│
├── .pytest_cache/
├── __pycache__/
└── source_files.py

🧪 Generated Tests

The generated_tests/ directory contains:

edge-case tests
boundary-condition tests
regression tests
integration-style tests

generated automatically for risky functions detected during analysis.

🕸 Knowledge Graph

knowledge_graph.json stores:

function relationships
import graph metadata
call graph structures
risk annotations
architectural insights

📝 Markdown Report

report.md contains:

static analysis findings
complexity hotspots
risky functions
dependency remarks
generated tests
execution summaries
failure diagnoses
refactoring suggestions

📊 Example Report Sections

The generated report includes:

⚠ Static analysis findings
📈 Complexity hotspots
🔴 Risky function rankings
🚧 Missing dependency diagnostics
🧪 Generated unit tests
🐛 Failure explanations
💡 Refactoring recommendations
🔁 Duplicate logic detection
🔗 Dependency insights

🤖 Supported LLM Providers

OpenAI-Compatible

RepoRelic supports any OpenAI-compatible endpoint:

OpenAI
DeepSeek
GPT-OSS
Claude-compatible gateways
Local inference servers
Ollama-compatible wrappers

Example custom base URL:

https://api.deepseek.com/v1

Gemini

RepoRelic also supports Google's Gemini models.

Get a free Gemini API key:

https://aistudio.google.com

🔒 Security

RepoRelic:

does NOT automatically execute unknown code without permission
requests execution approval before running generated tests
stores API keys locally inside VS Code settings
keeps repository analysis local except for LLM prompts
never modifies repository source files automatically

🚀 Performance Notes

Best performance is achieved with:

Python 3.11+
virtual environments (.venv)
repositories under ~5k Python files
modern LLM models

Recommended:

Apple Silicon / modern CPUs
SSD storage
isolated Python environments

🛠 Troubleshooting

Python Version Errors

Verify Python version:

python3 --version

Recommended:

Python 3.11+

Missing Dependencies

Install required packages:

pip install -r engine/requirements.txt

LLM Errors

Verify:

API key is valid
internet connection works
base URL is correct
provider setting matches API type

Test Execution Failures

Check:

project virtual environment
missing packages
invalid imports
incompatible Python versions
unsupported dependencies

Virtual Environment Issues

Recommended setup:

python3.11 -m venv .venv
source .venv/bin/activate

Then install:

pip install -r engine/requirements.txt

💻 Recommended Setup

macOS / Linux

python3.11 -m venv .venv
source .venv/bin/activate
pip install -r engine/requirements.txt

Windows

python -m venv .venv
.venv\Scripts\activate
pip install -r engine/requirements.txt

📸 Screenshots

Analysis Pipeline

Pipeline

Generated Markdown Report

Report

Generated Output Structure

Output

🧩 Extension Commands

Command	Description
`Analyze with RepoRelic`	Runs the complete autonomous analysis pipeline
`Open RepoRelic Report`	Opens generated markdown report
`Show RepoRelic Logs`	Displays execution and engine logs

🌟 Why RepoRelic?

RepoRelic combines:

static analysis
graph intelligence
LLM reasoning
automated testing
architectural diagnostics

into a single autonomous developer workflow inside VS Code.

It is designed for:

large repositories
rapid debugging
autonomous test generation
repository intelligence
engineering productivity

🛣 Roadmap

Planned features:

Multi-language support
Autonomous bug fixing
Interactive graph visualization UI
Semantic repository search
CI/CD integration
Incremental analysis
AI-generated patches
Repository embeddings
Multi-agent debugging
Local offline LLM support

🤝 Contributing

Contributions are welcome.

Ideas:

new analyzers
graph algorithms
improved prompts
execution sandboxing
visualization systems
performance optimizations

📄 License

MIT License

🔍 RepoRelic

AI-powered repository intelligence inside VS Code.

RepoRelic

YAPrepo