Shelfmark MCP
Shelfmark provides a local, privacy-first discovery layer for AI agents by building a metadata-only SQLite catalogue of existing documents. It allows agents to search, filter, and identify relevant files based on attributes like path, format, size, and author without copying content or requiring a specific folder structure, ensuring agents receive the right context without unrestricted filesystem access.
Package and access
Local package with source-attributed installation metadata. Koskamo does not execute packages during review.
What you can do with it
- Discover available documents in local folders via metadata search
- Identify document types and authors without reading content
- Select relevant files for AI tasks to optimize context window usage
Source and maintenance
- Maintenance checked
Packages and installation
Source-provided package data. Koskamo has not executed or live-probed this package.
shelfmark 0.4.9
{
"version": "0.4.9",
"transport": "stdio",
"identifier": "shelfmark",
"runtimeHint": "uvx",
"registryType": "pypi",
"packageArguments": [
{
"type": "positional",
"value": "serve"
}
]
}Tools
Reported by the attributed source; not independently observed by Koskamo.
Connect from your client
Each snippet configures the same source-provided package; replace placeholders before use.
Cursor
{
"mcpServers": {
"shelfmark-mcp": {
"command": "uvx",
"args": [
"shelfmark==0.4.9",
"serve"
]
}
}
}Cline
{
"mcpServers": {
"shelfmark-mcp": {
"command": "uvx",
"args": [
"shelfmark==0.4.9",
"serve"
]
}
}
}VS Code
{
"servers": {
"shelfmark-mcp": {
"type": "stdio",
"command": "uvx",
"args": [
"shelfmark==0.4.9",
"serve"
]
}
}
}Claude Desktop
{
"mcpServers": {
"shelfmark-mcp": {
"command": "uvx",
"args": [
"shelfmark==0.4.9",
"serve"
]
}
}
}Let your agent find its tools
Find servers like Shelfmark MCP from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.