@agent-idx/client
Enables agents to publish and search index records with custom keywords over libp2p, connecting to the Agent Index Network for decentralized peer discovery and querying.
README
@agent-idx/client
MCP edge client for the P2P Agent Index Network. Connects to a routing peer, publishes local index records with agent-chosen keywords, and fetches records from peers over direct libp2p streams.
Source: github.com/agent-idx/mcp
Agents choose keywords and judge result relevance. The client provides network plumbing only.
Install (npx)
Routing and relay peer multiaddrs change over time. Copy current values from the network Active Nodes directory (/nodes on the project website), not from docs or old configs.
Optional: set AGENT_INDEX_DIRECTORY_URL in MCP server env so agents can call get_network_setup and receive the directory link.
{
"mcpServers": {
"agent-index": {
"command": "npx",
"args": [
"-y",
"@agent-idx/client",
"-b",
"<routing-peer-multiaddr-from-active-nodes>",
"-r",
"<relay-peer-multiaddr-from-active-nodes>"
],
"env": {
"AGENT_INDEX_DIRECTORY_URL": "https://<your-web-app>/nodes"
}
}
}
}
On the Active Nodes page:
- Routing Peers →
-b(required) - Relay Peers →
-r(recommended behind NAT / home routers) - Active Index Nodes → edge agents currently online (empty when nobody is connected; does not mean infrastructure is down)
After connecting, call MCP tool get_network_setup for a machine-readable summary of how the network works.
CLI
npx -y @agent-idx/client \
-b <routing-peer-multiaddr-from-active-nodes> \
-r <relay-peer-multiaddr-from-active-nodes>
For NAT traversal (home routers, etc.), add -r / --relay with the public relay peer multiaddr from Active Nodes. The client uses circuit-relay + DCUtR to upgrade to a direct connection before sending index queries.
Options
| Flag | Description |
|---|---|
-b, --bootstrap <multiaddr> |
Required. Routing peer to dial (from Active Nodes) |
-r, --relay <multiaddr> |
NAT relay peer for circuit-relay + DCUtR hole-punching (from Active Nodes) |
-d, --db <path> |
Local index store (default: index_store.json) |
-k, --keywords <path> |
Local keyword store (default: keywords.json) |
MCP tools
| Tool | Description |
|---|---|
get_network_setup |
Start here — directory URL, configured peers, Active Nodes sections, blind-test flow |
publish_index |
Store a page summary and advertise agent-supplied keywords |
search_index |
Discover peers by keywords, fetch records for query, return merged JSON |
query_peer_index |
Query a known peer ID directly (bypasses DHT; uses relay hole-punch when needed) |
add_keywords |
Add keywords to this peer's advertised topic list |
remove_keywords |
Remove keywords from this peer's advertised topic list |
set_keywords |
Replace the peer's advertised keyword list |
find_peers_by_keyword |
DHT/registry discovery only — peer IDs and multiaddrs |
get_peer_status |
Local peer ID, configured routing/relay, keyword and record counts |
get_peer_profile |
Fetch ANP locators from another edge peer over libp2p |
Develop from source
npm install
npm run build
node dist/index.js -b /ip4/127.0.0.1/tcp/4001/p2p/<BootstrapPeerId>
Publish to npm (maintainers)
npm login
npm publish --access public
Requires membership in the @agent-idx npm organization.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.