yacy-searchLABSCO SUMMARY
Search that does not go through anyone's API.
The query never leaves your own infrastructure: your node answers it, from its own crawl or from peers. That is the trade to weigh — coverage depends on what your node and the network have indexed, and in exchange nobody is logging what you searched for.
What it is
A search call against a YaCy instance you run. YaCy is a peer-to-peer search engine, so the results come from your own index or from the peer network rather than from a commercial search API.
What you get
- Web search answered by your own YaCy node
- A choice per query between your local index and the global peer network
- A result limit you set, defaulting to ten
Requirements
A running YaCy instance — the server expects one at localhost:8090 unless you point it elsewhere. Launched with uvx on your PATH.
Setup effort
One command — uvx yacy-mcp
USE CASES
01
Searching without a search API key
Give an assistant web search backed by your own node, so no per-query billing or key management stands between a question and an answer.
02
Searching a corpus you crawled yourself
Restrict the query to the local index when you have crawled a documentation set or an intranet and want answers only from it.
03
Keeping queries off commercial search infrastructure
Route the search through the peer network instead of a provider that records it — the reason to run a YaCy node in the first place.
1 tool
Listed in project docs. Your client may show a different list.
BEFORE YOU INSTALL
What to know first
Tools you get
1 tool named in the project's documentation.
Requires
uvx on PATH
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
Server setup
TERMINAL
uvx yacy-mcpMCP Clients
Claude Desktop
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"yacy-mcp": {
"command": "uvx",
"args": [
"yacy-mcp"
],
"env": {
"YACY_URL": "http://localhost:8090"
}
}
}
}
