Baloosearch MCP Server
The Baloosearch MCP Server enables AI assistants to perform advanced file searches using KDE's Baloo search system.
Search Files (search_files tool):
Semantic searching - Search for terms within files across your system
Advanced query syntax - Use boolean operators (AND, OR, NOT), phrase searches with quotes, wildcards (
*), and grouping with parenthesesProperty-based filtering - Search by file metadata including:
General properties: filename, modified date, rating, tags, comments
Audio properties: Artist, Album, Genre, Duration, BitRate
Document properties: Author, Title, PageCount, WordCount
Image properties: dimensions, camera settings, GPS coordinates
Property comparisons - Use operators like
rating>3ormodified>2024-01-01File type filtering - Filter by Audio, Video, Image, Document (including Spreadsheet/Presentation), Archive, Text, or Folder
Configurable parameters - Set result limits, offsets, and restrict searches to specific directories
Show File Words (show_file_words tool):
View all indexed words and metadata from a specific file using its absolute path
Verify what content Baloo has extracted and indexed
Technical Features:
Returns results in structured JSON format
Compatible with any MCP client like Claude Desktop
Can be tested via command line using JSON-RPC messages
Works on any system with Node.js and KDE's baloosearch tool
Provides file search functionality using KDE's baloosearch tool, enabling semantic search across the file system to find relevant documents and content with configurable parameters like file type, directory scope, and result limits.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Baloosearch MCP Server find all PDF documents about project planning from last month"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Baloosearch MCP Server
A Model Context Protocol (MCP) server that provides file search functionality using KDE's baloosearch tool.
Description
This MCP server exposes a tool that allows AI assistants to search for terms in files using the KDE baloosearch utility. It enables semantic search across your file system to find relevant documents and content.
The server implements the Model Context Protocol specification and can be used with any MCP-compatible client, such as Claude Desktop or other AI assistants that support the protocol.
Related MCP server: fd-mcp
Features
Search for terms in files using KDE baloosearch
Configurable search parameters (limit, offset, directory, file type)
Advanced query syntax support (AND, OR, NOT, phrases, wildcards)
Property-based searches (Artist, Author, etc.)
File type filtering (Audio, Document, Image, etc.)
JSON output of search results
Full MCP compliance for integration with AI assistants
Prerequisites
Node.js (v18 or higher)
KDE baloosearch tool (typically installed with KDE desktop environment)
Installation
npm installRunning with npx
You can run this MCP server directly using npx without installing it globally:
npx baloosearch-mcpThis will start the MCP server that can be used with any MCP-compatible AI assistant.
Usage
Starting the Server
npm startThe server will start and listen for MCP requests via stdio.
Testing the Baloosearch Tool
npm run test-toolThis runs a direct test of the baloosearch functionality.
Tools
search_files
Search for terms in files using KDE baloosearch.
Query Syntax Examples:
Simple search:
"project plan"Multiple terms:
"budget AND marketing"(finds files with both terms)OR search:
"report OR presentation"(finds files with either term)Phrase search:
"\"strategic plan\""(finds exact phrase)Exclusion:
"financial -tax"(finds files with "financial" but not "tax")Wildcard:
"report*"(finds files with words starting with "report")Property search:
"Artist:\"Coldplay\""(finds audio files by artist)File type search:
"type:Audio"(finds audio files)Combined expressions:
"(type:Audio AND Artist:\"Coldplay\") OR (type:Document AND subject:\"music\")"Property comparisons:
"rating>3","modified>2024-01-01"
Supported File Types:
"Archive"(zip, tar, etc.)"Folder"(directories)"Audio"(mp3, wav, etc.)"Video"(mp4, avi, etc.)"Image"(jpg, png, etc.)"Document"(pdf, doc, etc.)"Spreadsheet"(xls, xlsx, etc.)"Presentation"(ppt, pptx, etc.)
"Text"(txt, etc.)
Common Properties for All Files:
filename(name of the file)modified(last modification date)mimetype(MIME type of file)tags(user-defined tags)rating(numeric rating 0-10)userComment(user comments)
Audio-Specific Properties:
Artist,Album,AlbumArtist,Composer,LyricistGenre,Duration,BitRate,Channels,SampleRateTrackNumber,ReleaseYear,Comment
Document-Specific Properties:
Author,Title,Subject,KeywordsPageCount,WordCount,LineCountLanguage,Copyright,PublisherCreationDate,Generator
Media-Specific Properties (Video/Images):
Width,Height,AspectRatio,FrameRate
Image-Specific Properties:
ImageMake,ImageModel,ImageDateTimePhotoFlash,PhotoFNumber,PhotoISOSpeedRatingsPhotoGpsLatitude,PhotoGpsLongitude,PhotoGpsAltitude
Parameters:
query(string, required): The search query terms. Supports advanced search syntax:AND: Requires both terms (e.g.,"budget AND marketing")OR: Requires either term (e.g.,"report OR presentation")NOTor-: Excludes terms (e.g.,"financial -tax"or"financial NOT tax")Phrase search: Use quotes for exact phrases (e.g.,
"\"project plan\"")Wildcards: Use
*for partial matching (e.g.,"report*")Property searches:
"Artist:\"Coldplay\""or"Author:\"Smith\""File type filters:
"type:Audio","type:Document","type:Image", etc.Property comparisons:
"rating>3","modified>2024-01-01"Grouping: Use parentheses to group terms (e.g.,
"(budget OR finance) AND 2024")
limit(number, optional): Maximum number of results to return (default: 10)offset(number, optional): Offset from which to start the search (default: 0)directory(string, optional): Limit search to specified directory (absolute path)type(string, optional): Type of data to be searched. Common types include:"Archive"(zip, tar, etc.)"Folder"(directories)"Audio"(mp3, wav, etc.)"Video"(mp4, avi, etc.)"Image"(jpg, png, etc.)"Document"(pdf, doc, etc.)"Spreadsheet"(xls, xlsx, etc.)"Presentation"(ppt, pptx, etc.)
"Text"(txt, etc.) Note: This parameter is an alternative to using "type:" in the query.
Returns:
JSON array of objects containing file paths that match the search criteria
Testing with Command Line
You can test the server directly from the command line using echo and pipes:
List Available Tools
echo '{"jsonrpc":"2.0","id":1,"method":"tools/list"}' | node src/server/mcp-server.jsSearch for Files
echo '{"jsonrpc":"2.0","id":2,"method":"tools/call","params":{"name":"search_files","arguments":{"query":"test","limit":3}}}' | node src/server/mcp-server.jsAdvanced Search Examples
# Search for files with both "budget" and "marketing"
echo '{"jsonrpc":"2.0","id":3,"method":"tools/call","params":{"name":"search_files","arguments":{"query":"budget AND marketing","limit":5}}}' | node src/server/mcp-server.js
# Search for audio files by a specific artist
echo '{"jsonrpc":"2.0","id":4,"method":"tools/call","params":{"name":"search_files","arguments":{"query":"type:Audio AND Artist:\"ArtistName\"","limit":5}}}' | node src/server/mcp-server.js
# Search for documents with high ratings
echo '{"jsonrpc":"2.0","id":5,"method":"tools/call","params":{"name":"search_files","arguments":{"query":"type:Document AND rating>3","limit":5}}}' | node src/server/mcp-server.js
# Search for recent documents
echo '{"jsonrpc":"2.0","id":6,"method":"tools/call","params":{"name":"search_files","arguments":{"query":"type:Document AND modified>2024-01-01","limit":5}}}' | node src/server/mcp-server.js
# Search for images with specific properties
echo '{"jsonrpc":"2.0","id":7,"method":"tools/call","params":{"name":"search_files","arguments":{"query":"type:Image AND PhotoFNumber<2.8","limit":5}}}' | node src/server/mcp-server.jsIntegration with Claude Desktop
To use this server with Claude Desktop:
Add the following to your
claude_desktop_config.json:
{
"mcpServers": {
"baloosearch": {
"command": "node",
"args": ["/path/to/baloosearch-mcp/src/server/mcp-server.js"],
"env": {}
}
}
}Restart Claude Desktop
The baloosearch tool will be available in the tools list
Project Structure
baloosearch-mcp/
├── src/
│ ├── server/
│ │ └── mcp-server.js # Main MCP server implementation
│ ├── tools/
│ │ └── baloosearch-tool.js # Baloosearch tool implementation
│ ├── test-tool.js # Test script for the tool
│ └── test-server.js # Test script for the server
├── test/
│ ├── integration.js # Integration test script
│ └── server-test.sh # Server test script
├── package.json # Project configuration
└── README.md # This fileDevelopment
Running Tests
# Test the baloosearch tool directly
npm run test-tool
# Test the MCP server startup
npm run test-serverPublishing to npm
To publish this package to npm:
Create an account at npmjs.com if you don't have one
Log in to npm:
npm loginPublish the package:
npm publish
After publishing, users will be able to run your MCP server using:
npx baloosearch-mcpModifying the Server
The main server implementation is in src/server/mcp-server.js. You can add additional tools or modify existing ones by following the same pattern as the search_files tool.
Modifying the Baloosearch Tool
The baloosearch tool implementation is in src/tools/baloosearch-tool.js. You can modify the search parameters or add additional functionality as needed.
Troubleshooting
"baloosearch: command not found"
Make sure you have KDE desktop environment installed with baloosearch available. You can test this by running:
which baloosearchServer exits immediately
This is normal behavior for MCP servers using stdio transport. The server waits for JSON-RPC messages via stdin/stdout.
No search results
Make sure your baloosearch index is up to date. You can update it by running:
balooctl monitorLicense
This project is licensed under the MIT License.
Available Tools
2 toolssearch_filesC
Search for terms in files using KDE baloosearch.
Query Syntax Examples:
Simple search: "project plan"
Multiple terms: "budget AND marketing" (finds files with both terms)
OR search: "report OR presentation" (finds files with either term)
Phrase search: ""strategic plan"" (finds exact phrase)
Exclusion: "financial -tax" (finds files with "financial" but not "tax")
Wildcard: "report*" (finds files with words starting with "report")
Property search: "Artist:"Coldplay"" (finds audio files by artist)
File type search: "type:Audio" (finds audio files)
Combined expressions: "(type:Audio AND Artist:"Coldplay") OR (type:Document AND subject:"music")"
Inputs: query (required), limit, offset, directory, +1 more.
show_file_wordsA
Show indexed content & meta-data in a file using KDE balooshow.
This tool retrieves all the indexed words & meta-data from a specific file that has been indexed by KDE's Baloo search system.
Inputs: path (required).
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 2 tools
The two tools have clearly distinct purposes: search_files searches across files for terms, while show_file_words displays indexed content and metadata for a specific file. There is no overlap or ambiguity between them.
Both tool names follow a consistent verb_noun pattern with snake_case: search_files and show_file_words. The naming is predictable and readable throughout.
With only 2 tools, the server feels thin for its apparent scope of file search and metadata display. A typical search system might include additional operations like listing indexed files, updating indexes, or filtering results, making this set under-scoped.
Inferring the domain as file search and indexing, there are significant gaps: no tools for managing the index (e.g., refresh, status), listing available files or properties, or performing advanced queries beyond basic search. This limits agent workflows and could lead to dead ends.
Maintenance
Related MCP Connectors
Securely search and manage workspace context files for AI agents and teams.
Read and search folders on your own computer with explicit consent. Files are never uploaded.
Provides AI assistants with access to Seltz's powerful Web Search capabilities.
Search photos and video moments by visual meaning, people, dates, and folders across local libraries, Apple Photos, and OneDrive. Manage collections, groups, favorites, people, and duplicate review. Synced metadata remains searchable while the desktop is offline; local and Apple Photos media access needs the owning device online with remote access enabled. OneDrive media can be accessed directly from the cloud. OAuth or Bearer API keys. Paid Cloud Access required; hosted visual search uses prepaid credits. Text-in-image search is semantic matching, not guaranteed exact OCR. Setup: https://organizewith.ai/en/guides/connect-ai-assistants
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceEnables intelligent file searching in local directories using natural language queries. Supports searching by file type, filename patterns, and content across multiple formats including PDF, Word, Excel, and text files with AI-powered relevance scoring.-
- AlicenseNot gradedqualityDmaintenanceProvides fast file search capabilities using fd (a modern find alternative), enabling AI assistants to efficiently navigate codebases, search file contents with ripgrep, and execute commands on matched files.1MIT
- AlicenseNot gradedqualityBmaintenanceProvides AI assistants with semantic search and read access to local files and directories, enabling knowledge retrieval from indexed content.36 npm17MIT
- AlicenseNot gradedqualityDmaintenanceEnables AI assistants to search local files on Windows using the Everything search engine, supporting basic and complex file searches with filters like date, size, media type, and document type.4MIT