SolSigs Python Client ·

Pay-per-call Solana data APIs for AI agents via x402 micropayments.

No API keys. No subscriptions. Agents pay in USDC per call.

Claude Desktop (Native x402)

Claude agents can pay for SolSigs APIs natively — 30 second setup.

pip install solsigs-py

Then add this to ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "solsigs": {
      "command": "solsigs-mcp"
    }
  }
}

Restart Claude Desktop. Your agent now has 15 Solana research tools with native x402 micropayments.

Why a local command? Claude Desktop only supports stdio transport. solsigs-mcp is a local proxy that bridges stdio ↔ the remote SolSigs MCP server over SSE. Full example: examples/claude_solsigs_agent.py

🔗 Anthropic official x402 docs

Install

pip install solsigs-py              # base client
pip install solsigs-py[langgraph]   # + LangGraph tools

Quick Start

from solsigs import SolSigsClient

client = SolSigsClient()

# DEX price for any Solana token
price = client.get_dex_price("DezXAZ8z7PnrnRJjz3wXBoRgixCa6xjnB7YaB1pPB263")
print(price)  # → $0.002 USDC

# Multi-token batch pricing + OHLCV
prices = client.batch_token_pricing(["SOL", "BONK", "WIF"], period="1h")

# Wallet risk scoring
report = client.score_wallet("7EcDhSYGxXyscszYEp35KHN8vvw3svAuLKTzXwCFLtV")

# Arbitrage scan
arbs = client.scan_arbitrage(min_spread_pct=0.5)

Payment flow

Agent calls endpoint → 402 Payment Required → Pays USDC → Retries → Gets data

The client handles the HTTP layer. Integrate your agent's Solana wallet for automated x402 payment.

MCP Server

SolSigs ships as an MCP server (Model Context Protocol) — Claude, Hermes, Cursor, and any MCP-compatible agent can call SolSigs tools natively.

Claude Desktop users: use the solsigs-mcp command above (stdio bridge).

Hermes, Cursor, and other SSE-capable MCP clients:

{
  "mcpServers": {
    "solsigs": {
      "url": "https://solsigs.com/sse",
      "transport": "sse"
    }
  }
}

15 MCP tools — all 14 HTTP endpoints + Polymarket prediction markets:

Data & Pricing (5)

Trading & Arbitrage (2)

Intelligence & Tracking (5)

Infrastructure (3)

LangGraph Integration

from solsigs.langgraph_tools import SOLSIGS_LANGGRAPH_TOOLS

# Add to your LangGraph agent
from langgraph.prebuilt import create_react_agent

agent = create_react_agent(model, SOLSIGS_LANGGRAPH_TOOLS)

# Your agent can now:
agent.invoke({"messages": ["What's the current price of BONK?"]})
agent.invoke({"messages": ["Compare staking APY across Solana protocols"]})

See examples/langgraph_agent.py for a full multi-step agent demo.

Available HTTP Endpoints (14)

Data & Pricing (5)

Trading & Arbitrage (2)

Intelligence & Tracking (5)

Infrastructure (2)

SolSigs ProofGuard

ProofGuard is a signed receipt and refund-evidence layer for x402-paid agent endpoints on Solana.

It returns:

• endpoint trust score;
• payment expectation receipt;
• payment fulfillment receipt;
• refund evidence package.

Live endpoint:

POST https://solsigs.com/proofguard/evaluate
Price: 0.003 USDC via x402 on Solana mainnet

ProofGuard explicitly labels evidence mode as solsigs_instrumented, proofguard_probed, or self_attested, so agents can distinguish real observed proof from caller-supplied claims.

See docs/proofguard.md.

Why x402?

• No API keys — payment is the only gate
• No subscriptions — pay only for what you use
• Agent-native — designed for autonomous agent commerce
• Solana speed — transactions finalize in <1s, fees <$0.001
• Claude native — Anthropic official x402 support, zero setup for Claude users

Links

• SolSigs Homepage
• Claude + SolSigs Example
• Anthropic x402 Docs
• x402 Protocol
• x402scan Explorer

License

MIT