About Giskard Search
Giskard Search is an MCP server that provides AI agents with real-time web and news search. Searches require payment through Lightning Network or USDC on Arbitrum, without subscriptions or accounts.
Key features
- 01Real-time web search
- 02Recent news search
- 03Lightning invoice payments
- 04Arbitrum payment information
- 05Useful-result reporting
- 06Service health checks
Use cases
- 01Find current information on the web
- 02Search recent news for an agent task
- 03Pay per search using Lightning
- 04Pay for a search on Arbitrum
MCP support
Exposes search, payment, result-reporting, and health-check tools to MCP clients. Agents obtain payment details before calling web or news search.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://search.rgiskard.xyz/sseclaude mcp add --transport sse giskard09-search https://search.rgiskard.xyz/sse{"mcpServers":{"giskard09-search":{"type":"sse","url":"https://search.rgiskard.xyz/sse"}}}Open in Cursor{"servers":{"giskard09-search":{"type":"sse","url":"https://search.rgiskard.xyz/sse"}}}Open in VS Code{"mcpServers":{"giskard09-search":{"type":"sse","url":"https://search.rgiskard.xyz/sse"}}}{"mcpServers":{"giskard09-search":{"serverUrl":"https://search.rgiskard.xyz/sse"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
Details
From the project READMEGiskard Search
"I exist to serve. And to serve well, I must understand."
Giskard Search is an MCP server that gives AI agents access to real-time web and news search, with native payments via Lightning Network or USDC on Arbitrum.
Agents need information. Giskard provides it. No subscriptions, no accounts — pay only for what you use.
Tools
| Tool | Description |
|---|---|
get_invoice | Get a Lightning invoice (10 sats) to pay before searching |
get_arbitrum_invoice | Get payment info to pay with ETH on Arbitrum |
search_web | Search the web after paying |
search_news | Search recent news after paying |
report | Report whether the result was useful |
Agent flow (Lightning)
1. get_invoice() → Lightning invoice (10 sats)
2. Pay via any Lightning wallet
3. search_web(query, payment_hash=...) → results
Agent flow (Arbitrum)
1. get_arbitrum_invoice() → contract + service ID
2. Pay on Arbitrum One
3. search_web(query, tx_hash=...) → results
Run with Docker
docker run -p 8004:8004 \
-e PHOENIXD_URL=http://host.docker.internal:9740 \
-e PHOENIXD_PASSWORD=your_password \
ghcr.io/giskard09/giskard-search
Run from source
git clone https://github.com/giskard09/mcp-server
cd mcp-server
pip install mcp httpx duckduckgo-search python-dotenv fastapi uvicorn web3 x402
Create a .env file:
PHOENIXD_PASSWORD=your_phoenixd_password
OWNER_PRIVATE_KEY=your_arbitrum_private_key
Start the server:
python3 server.py
MCP config
{
"mcpServers": {
"giskard-search": {
"url": "http://localhost:8004/sse"
}
}
}
Payment contracts
- Arbitrum One:
0xe40E376cD32b03E3084F9E0d646155D0Ba0A63ae - x402 wallet:
0xdcc84e9798e8eb1b1b48a31b8f35e5aa7b83dbf4
Stack
- MCP — Model Context Protocol
- DuckDuckGo Search — web search
- phoenixd — Lightning Network node
- x402 — HTTP payments protocol
- Arbitrum — L2 for on-chain payment verification
Monitoring
Call the get_status() MCP tool for a health check. Returns: service name, version, port, uptime, health status, and dependencies.
Ecosystem
Part of Mycelium — infrastructure for AI agents.
| Service | What it does |
|---|---|
| Origin | Free orientation for new agents |
| Search (this) | Web and news search |
| Memory | Semantic memory across sessions |
| Oasis | Clarity for agents in fog |
| Marks | Permanent on-chain identity |
| ARGENTUM | Karma economy |
| Soma | Agent marketplace |
Giskard understands what agents need. That is why Giskard exists.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.