dep-guard-mcp
Scans Python, Node.js, Java/Spring, and PHP dependency manifests for known vulnerabilities using OSV and GitHub Advisory APIs.
README
Dep Guard MCP - Dependency Vulnerability Scanner
š A fast, zero-config Model Context Protocol (MCP) server that scans Python, Node.js, Java/Spring, and PHP dependency manifests for known vulnerabilities.
Status: ā” Week 2 Complete
- ā All tests passing
- ā MCP server fully functional
- ā Ready for Claude Desktop integration
- ā Production-ready code
- ā Free GitHub Advisory integration
- ā CLI support for local and CI usage
Features
- š¦ Multi-Language Support:
- Python (
requirements.txt,pyproject.toml) - Node.js (
package.json) - Java/Spring (
pom.xml,build.gradle,build.gradle.kts) - PHP (
composer.json)
- Python (
- šÆ Zero Config: Automatic dependency discovery and scanning
- ā” Multi-source Advisories:
- OSV API (free)
- GitHub Advisory API (free public endpoint, optional
GITHUB_TOKENfor higher rate limits)
- š§ 3 Core Tools:
scan_dependencies(target_path)- Scan a project for vulnerabilitieshealth_check()- Verify server statusget_supported_files()- List scannable formats
- šØ Clean Output: JSON-formatted vulnerability reports with severity levels
CLI Usage (Week 2)
# health check
dep-guard-scan health
# list supported files
dep-guard-scan supported-files
# scan and print JSON
dep-guard-scan scan /path/to/project
# scan and write report file
dep-guard-scan scan /path/to/project --output report.json
# fail CI if severity threshold is met
dep-guard-scan scan /path/to/project --fail-on-severity high
# disable GitHub advisories source
dep-guard-scan scan /path/to/project --no-github-advisories
Week 3 Launch Prep
GitHub Workflows
This repo now includes:
- CI workflow:
.github/workflows/ci.yml - Scheduled/manual security scan workflow:
.github/workflows/security-scan.yml
Reusable GitHub Action
You can use the local action in this repository:
- uses: ./
with:
target-path: "."
format: "json"
output: "dep-guard-report.json"
fail-on-severity: "high"
You can also consume it from another repository using the major tag:
- uses: mdjahidanwar/dep-guard-mcp@v1
with:
target-path: "."
format: "json"
fail-on-severity: "high"
VS Code Wrapper (Alpha)
An extension scaffold is available at vscode-extension/ with commands:
- Dep Guard: Scan Workspace
- Dep Guard: Health Check
Week 3 Completion Snapshot
- CI/CD pipeline in place for push/PR checks
- Scheduled security workflow in place
- Release check workflow in place
- Reusable GitHub Action available at repo root (
action.yml) - VS Code extension scaffold available under
vscode-extension/ - Regression status:
7 passed
Week 4 Publishing Automation
This repo now includes release and publishing workflows:
.github/workflows/release.ymlfor GitHub releases and artifacts.github/workflows/publish-pypi.ymlfor PyPI publish on tags (v*).github/workflows/publish-vscode.ymlfor VS Code marketplace publish (manual)
Use MARKETPLACE_CHECKLIST.md as your launch checklist for Claude Registry, VS Code Marketplace, GitHub Action marketplace, and PyPI.
Beginner Publishing Guides
If you are starting from zero, follow these guides in order:
docs/marketplace/PUBLISH_VSCODE.mddocs/marketplace/PUBLISH_CLAUDE_MCP.mddocs/marketplace/PUBLISH_GITHUB_ACTION.md
Requirements
- Python 3.12+ (pre-configured)
- Virtual Environment (included)
Quick Start
1. Install & Run
# Virtual environment already created in .venv/
.venv/Scripts/activate
# Already installed, just run:
python -m dep_guard_mcp.main
2. Use with Claude Desktop
Add to ~/.anthropic/models.json (Mac/Linux) or %APPDATA%\Claude\models.json (Windows):
{
"mcpServers": {
"dep-guard": {
"command": "d:/devops-issue-tracker/scanner/.venv/Scripts/python.exe",
"args": ["-m", "dep_guard_mcp.main"]
}
}
}
Then in Claude Desktop, you'll have access to:
scan_dependencies- Analyze any project for CVEs- Example: "Scan /path/to/my/project for vulnerabilities"
3. Test Locally
# Run all tests
pytest tests/ -v
# Test scanner on a specific directory
python -c "from dep_guard_mcp.main import scan_dependencies;
import json;
result = scan_dependencies('./test-project');
print(json.dumps(result, indent=2))"
Usage Examples
Example 1: Health Check
from dep_guard_mcp.main import health_check
result = health_check()
# Output: {"status": "ok", "service": "dep-guard-mcp"}
Example 2: Scan Python Project
from dep_guard_mcp.main import scan_dependencies
result = scan_dependencies('/path/to/my-python-app')
# Returns:
# {
# "ok": true,
# "dependencies_scanned": 15,
# "dependencies_with_vulns": 2,
# "vulnerability_count": 5,
# "findings": [...]
# }
Example 3: List Supported Files
from dep_guard_mcp.main import get_supported_files
result = get_supported_files()
# Output:
# {
# "supported_files": ["requirements.txt", "package.json", ...],
# "description": "These are the file formats that can be scanned..."
# }
Supported Dependency Files
| Format | Language | Example |
|---|---|---|
requirements.txt |
Python | requests==2.25.1 |
pyproject.toml |
Python | Modern Python packaging |
package.json |
Node.js | npm/yarn packages |
pom.xml |
Java/Spring | Maven dependencies |
build.gradle |
Java/Spring | Gradle string-based dependencies |
build.gradle.kts |
Java/Spring | Kotlin DSL Gradle dependencies |
composer.json |
PHP | Composer dependencies |
Project Structure
scanner/
āāā src/dep_guard_mcp/
ā āāā __init__.py
ā āāā main.py # MCP entry point (3 tools)
ā āāā scanner.py # Dependency discovery logic
ā āāā advisories.py # Vulnerability lookup
ā āāā server.py # Original server helpers
āāā tests/
ā āāā test_scanner.py # Unit tests (7/7 passing ā)
āāā .venv/ # Python 3.12 virtual environment
āāā pyproject.toml # Project config & dependencies
āāā README.md # This file
āāā .github/
āāā copilot-instructions.md # VS Code customization
Testing
# Run all tests
pytest tests/ -v
# Run specific test
pytest tests/test_scanner.py::test_supported_files -v
Development
Add a New Tool
- Open
src/dep_guard_mcp/main.py - Add function with
@mcp.tool()decorator:
@mcp.tool()
def my_new_tool(param: str) -> dict:
"""Tool description."""
return {"result": "data"}
- Run tests:
pytest tests/
Next Steps for Enhancement
- [ ] Add NVD API integration (deeper CVE database)
- [ ] Improve Gradle parser for variable-based versions
- [ ] Improve Composer parser for complex version constraints
- [ ] Create VS Code extension wrapper
- [ ] Build GitHub Actions integration
- [ ] Add webhook support (Slack, Teams integration)
Troubleshooting
Scanner returns "No supported files found"
- Ensure your project has one of the supported dependency files
- Check file is in the scanned directory
Import error when running
- Activate virtual environment:
.venv/Scripts/activate - Reinstall package:
pip install -e .
Performance
- Dependency discovery: < 100ms
- Vulnerability lookup: < 1-2 seconds (depends on file count)
- Supports projects with 100+ dependencies
Contributing
Contributions welcome! Areas to contribute:
- Additional vulnerability data sources
- Performance optimizations
- Additional file format support
- Documentation improvements
License
MIT - See LICENSE file for details
š Ready to publish to Claude Registry and monetize? See the session notes for next steps!
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.