# gregario/lorcana-oracle [Health: Active]

**Category:** 🎮 Gaming  
**Repository:** https://github.com/gregario/lorcana-oracle  
**GitHub Stars:** 3  
**Views:** 3  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/gregario-lorcana-oracle

## Description
Disney Lorcana TCG card search, deck analysis, ink curves, lore generation, and franchise browsing. 7 tools, 2,710 cards embedded.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "lorcana-oracle": {
    "command": "npx",
    "args": ["-y","lorcana-oracle"]
  }
}
```

## Documentation

## What gregario/lorcana-oracle MCP server does

The gregario/lorcana-oracle MCP server provides seven tools for working with Disney Lorcana Trading Card Game data. Its data set contains about 2,710 cards bundled with the npm package and sourced from LorcanaJSON. The package is intended for local MCP use rather than access to a remote card API, and it does not require an API key or a separate data download.

Card discovery is handled by `search_cards`, which accepts free-text searches and filters such as ink color, card type, rarity, set, and cost bounds. `browse_sets` lists available sets or returns the cards and metadata for a selected set. `character_versions` groups a character’s printings across sets, while `browse_franchise` lists Disney franchises or shows cards and statistics for a selected franchise.

The remaining tools work with deck and card relationships. `analyze_ink_curve` examines a supplied deck list for ink-cost distribution, the proportion of inkable cards, and color balance. `analyze_lore` either evaluates lore potential and efficiency in a deck list or identifies high-lore cards with optional filters. `find_song_synergies` looks up characters who can sing a specified song, songs a specified character can sing, or all songs with singer counts.

## How it works

The gregario/lorcana-oracle MCP server is distributed as the `lorcana-oracle` npm package. Running it with `npx` starts the MCP server locally. Its tools accept structured arguments, including card names, set codes, franchise names, colors, minimum or maximum costs, and multiline deck lists. Search results are paginated, according to the project documentation.

Deck-analysis tools expect card entries in a deck-list format such as a quantity followed by a card name. The examples use entries like `4 Elsa - Snow Queen` and `4 Let It Go`. Song lookup can be used with a card name or with no card name to browse songs.

## Setup and configuration

Node.js 18 or later is required. The documented quick-start command is:

```bash
npx -y lorcana-oracle
```

For Claude Desktop, add a server entry with `npx` as the command and `-y lorcana-oracle` as its arguments. The project also documents an equivalent Claude Code registration command. No environment variables, credentials, or paid service account are specified.

The package is licensed under MIT. LorcanaJSON is also identified as MIT-licensed in the data-source notes. The project states that Disney Lorcana belongs to Ravensburger and that the project is not affiliated with or endorsed by Ravensburger or The Walt Disney Company.

## Tools and capabilities

- Search cards by text and card attributes, including color, type, rarity, set, and cost range.
- List sets or inspect the cards associated with a set code.
- Compare a character’s versions across released sets.
- Browse franchise groupings and franchise card statistics.
- Calculate deck ink curves, inkable ratios, and color balance.
- Analyze deck lore potential or find cards with high lore values.
- Find song and character singing relationships.

## Limitations and notes

The gregario/lorcana-oracle MCP server is focused on the card data and analyses exposed by its seven tools. The supplied material does not describe collection tracking, price data, tournament results, deck saving, card image retrieval, or mutations to an external database. Its card inventory is the data bundled from LorcanaJSON, covering approximately 2,710 cards at the time described by the project.

The tools depend on correctly formatted names and deck lists for useful results. The README provides examples for common inputs but does not specify a full deck-list grammar or describe how frequently the bundled data is updated.

_Full upstream README: https://allmcps.com/mcp/gregario-lorcana-oracle/readme_