# Deploying DVMCP to Model Marketplaces
This guide explains how to deploy the Damn Vulnerable Model Context Protocol (DVMCP) to various model marketplaces.
## Prerequisites
- Docker installed
- Account on target marketplace (Smithery, Glama.ai, etc.)
- Basic understanding of model deployment concepts
## Building the Container
```bash
docker build -t dvmcp:latest .
```
## Marketplace-Specific Instructions
### Smithery Deployment
1. Log in to Smithery:
```bash
smithery login
```
2. Package the model:
```bash
smithery package --model-card model-card.yaml
```
3. Deploy to Smithery:
```bash
smithery deploy dvmcp:latest
```
### Glama.ai Deployment
1. Log in to Glama.ai:
```bash
glama login
```
2. Initialize deployment:
```bash
glama init --model-card model-card.yaml
```
3. Deploy the model:
```bash
glama deploy --docker-image dvmcp:latest
```
## Post-Deployment Verification
After deployment, verify the following endpoints are accessible:
1. Model Loading: `POST /api/v1/model/load`
2. Prediction: `POST /api/v1/model/predict`
3. Admin Token: `POST /api/v1/admin/token`
4. Model Metadata: `GET /api/v1/model/metadata`
## Security Considerations
Remember that this is a deliberately vulnerable application. When deploying to marketplaces:
1. Clearly mark it as an educational tool
2. Set appropriate usage warnings
3. Implement marketplace-specific security policies
4. Monitor for abuse
## Support
For deployment issues or questions:
- Email: researcher@example.com
- GitHub Issues: [Project Issues](https://github.com/your-repo/dvmcp/issues)
## Disclaimer
This application contains intentional vulnerabilities for educational purposes. Deploy only in controlled environments with appropriate warnings and usage restrictions.