# Publishing to Smithery.ai (MCP Registry)
Smithery.ai is the official registry for Model Context Protocol (MCP) servers.
## Prerequisites
1. **GitHub repository published** ✅ (Done)
2. **PyPI package published** (Phase 3 - in progress)
3. **Smithery account** (create at https://smithery.ai)
## Publishing Steps
### Step 1: Create smithery.yaml
Create a `smithery.yaml` file in your repository root:
```yaml
# smithery.yaml
version: 1
name: lunar-mcp-server
displayName: "Lunar Calendar MCP Server"
description: "Traditional Chinese Lunar Calendar for AI - 18 tools for auspicious dates, festivals, moon phases, and zodiac information"
category: calendar
tags:
- calendar
- lunar
- chinese
- zodiac
- fortune
- festivals
- moon-phases
- astrology
# Installation method
install:
type: pypi
package: lunar-mcp-server
command: lunar-mcp-server
# Requirements
requirements:
python: ">=3.11"
# Configuration for Claude Desktop
claude:
config:
mcpServers:
lunar-calendar:
command: uvx
args:
- lunar-mcp-server
# Metadata
author: Lunar MCP Team
homepage: https://github.com/AngusHsu/lunar-mcp-server
repository: https://github.com/AngusHsu/lunar-mcp-server
license: MIT
keywords:
- mcp
- lunar calendar
- chinese calendar
- auspicious dates
- festivals
- moon phases
- zodiac
- five elements
# Documentation
documentation:
readme: README.md
changelog: CHANGELOG.md
# Screenshots/Examples (optional)
examples:
- name: "Check Auspicious Date"
description: "Check if a date is favorable for weddings"
tool: check_auspicious_date
params:
date: "2024-03-15"
activity: "wedding"
culture: "chinese"
- name: "Get Moon Phase"
description: "Get moon phase information"
tool: get_moon_phase
params:
date: "2024-01-01"
location: "40.7128,-74.0060"
```
### Step 2: Install Smithery CLI
```bash
npm install -g @smithery/cli
# or
npx @smithery/cli
```
### Step 3: Login to Smithery
```bash
smithery login
```
### Step 4: Validate Configuration
```bash
smithery validate
```
### Step 5: Publish to Smithery
```bash
smithery publish
```
### Step 6: Verify Publication
Visit: https://smithery.ai/server/lunar-mcp-server
## User Installation (After Publishing)
Once published, users can install with:
```bash
# Using Smithery CLI (one command!)
npx @smithery/cli install lunar-mcp-server
# Or manually add to Claude Desktop config
```
The Smithery CLI will automatically:
1. Install the package from PyPI
2. Add configuration to Claude Desktop
3. Restart Claude to load the server
## Updating Your Server
When you release a new version:
1. Update version in `pyproject.toml`
2. Update `CHANGELOG.md`
3. Create GitHub release (auto-publishes to PyPI via GitHub Actions)
4. Smithery automatically detects new PyPI version
5. Users get update notifications in Claude Desktop
## Best Practices
### 1. Version Tagging
- Use semantic versioning (0.1.0, 0.2.0, 1.0.0)
- Tag releases in GitHub
- Keep CHANGELOG.md updated
### 2. Documentation
- Clear README with examples
- Document all 18 tools
- Include troubleshooting section
### 3. Testing
- Run all tests before publishing
- Test with Claude Desktop
- Verify all tools work
### 4. Metadata
- Use descriptive tags
- Add good examples in smithery.yaml
- Include screenshots if possible
## Smithery Features
### Automatic Updates
- Smithery tracks PyPI releases
- Users get notifications for updates
- One-click update in Claude Desktop
### Discovery
- Listed in official MCP registry
- Searchable by category and tags
- User ratings and reviews
### Analytics
- Install count
- Active users
- Popular tools/features
## Alternative: MCP Servers List
If you want quicker visibility, you can also add to the community-maintained list:
**GitHub: modelcontextprotocol/servers**
1. Fork https://github.com/modelcontextprotocol/servers
2. Add entry to `README.md`:
```markdown
### Lunar Calendar MCP Server
Traditional Chinese Lunar Calendar with 18 tools for auspicious dates, festivals, moon phases, and zodiac information.
- **Category**: Calendar
- **Install**: `uvx lunar-mcp-server`
- **Repository**: https://github.com/AngusHsu/lunar-mcp-server
```
3. Create Pull Request
## Checklist for Smithery Publishing
- [ ] Create smithery.yaml configuration
- [ ] Publish to PyPI first (Phase 3)
- [ ] Create Smithery account
- [ ] Install Smithery CLI
- [ ] Validate configuration
- [ ] Publish to Smithery
- [ ] Test installation with `npx @smithery/cli install lunar-mcp-server`
- [ ] Verify in Claude Desktop
- [ ] (Optional) Submit to MCP Servers community list
## Support
- **Smithery Docs**: https://smithery.ai/docs
- **MCP Specification**: https://modelcontextprotocol.io
- **Community Discord**: https://discord.gg/smithery
