# Enhanced Wikidata MCP Server with Authentication

This is an enhanced version of the Wikidata MCP server that supports API token authentication for increased rate limits.

## Features

- ✅ **API Token Authentication** - 5,000 requests/hour (vs 500 without token)
- ✅ **Wikimedia User-Agent Policy Compliant**
- ✅ **All standard Wikidata tools** (search, SPARQL, metadata)
- ✅ **Heritage Institution Identifiers** - Get ISIL, VIAF, GND codes

## Rate Limits

| Authentication | Rate Limit |
|----------------|-----------|
| No token | 500 requests/hour |
| Personal API token | 5,000 requests/hour |
| Bot account | Unlimited |

## Setup

### 1. Get API Token

1. Visit: https://api.wikimedia.org/wiki/Special:AppManagement
2. Click "Create a personal API token"
3. Fill in:
   - **Name**: `GLAM Data Extraction`
   - **Description**: `Heritage institution data enrichment`
   - **Grants**: Select "Read" permissions
4. Copy the token

### 2. Configure Environment

Add to `/Users/kempersc/apps/glam/.env`:

```bash
WIKIDATA_API_TOKEN=your_actual_token_here
WIKIMEDIA_CONTACT_EMAIL=your@email.com
```

### 3. Install Dependencies

```bash
cd /Users/kempersc/apps/glam/mcp_servers/wikidata_auth
uv venv
source .venv/bin/activate  # On macOS/Linux
uv pip install -e .
```

### 4. Configure Claude Desktop

Add to `~/Library/Application Support/Claude/claude_desktop_config.json`:

```json
{
  "mcpServers": {
    "wikidata-auth": {
      "command": "uv",
      "args": [
        "run",
        "--directory",
        "/Users/kempersc/apps/glam/mcp_servers/wikidata_auth",
        "python",
        "server.py"
      ],
      "env": {
        "WIKIDATA_API_TOKEN": "${WIKIDATA_API_TOKEN}",
        "WIKIMEDIA_CONTACT_EMAIL": "${WIKIMEDIA_CONTACT_EMAIL}"
      }
    }
  }
}
```

## Available Tools

### `search_entity(query: str)`
Search for Wikidata entity IDs.

Example: `search_entity("Rijksmuseum")` → `"Q190804"`

### `search_property(query: str)`
Search for Wikidata property IDs.

Example: `search_property("ISIL code")` → `"P791"`

### `get_properties(entity_id: str)`
Get all properties for an entity.

Example: `get_properties("Q190804")` → `["P31", "P17", "P131", ...]`

### `get_metadata(entity_id: str, language: str = "en")`
Get label and description.

Example: `get_metadata("Q190804", "en")` → `{"Label": "Rijksmuseum", "Description": "national museum in Amsterdam"}`

### `get_identifiers(entity_id: str)`
Get external identifiers (ISIL, VIAF, etc.).

Example: `get_identifiers("Q190804")` → `{"ISIL": "NL-AmRMA", "VIAF": "148691498", ...}`

### `execute_sparql(sparql_query: str)`
Execute SPARQL queries.

Example:
```sparql
SELECT ?item ?itemLabel WHERE {
  ?item wdt:P791 "NL-AmRMA" .
  SERVICE wikibase:label { bd:serviceParam wikibase:language "en" }
}
```

## Usage in GLAM Project

This MCP is designed for the GLAM Data Extraction project to:
- Enrich heritage institution records with Wikidata identifiers
- Resolve GHCID collisions using Q-numbers
- Validate institution names and locations
- Cross-reference with authoritative identifiers (ISIL, VIAF)

## License

MIT License (based on https://github.com/zzaebok/mcp-wikidata)