The full upstream README, mirrored here for reference. Install config, tool schemas, adoption signals, and an original overview live on the AlphaBotPro Trading API listing page.
A minimal, runnable example showing an AI agent discovering, paying for, and unlocking live trading data over MCP using the x402 machine-payment pattern — no human, no dashboard, no checkout form.
This talks to the real, live AlphaBotPro MCP server. Nothing here is a mock.
initialize — standard MCP JSON-RPC handshaketools/list — discover the available toolsget_catalog — a public tool: prices and sample payloads, no authget_signals without a key — a clean, structured payment_required
error (the x402 pattern) instead of a silent failurepurchase_access — request a crypto invoice (USDT/USDC on TRON,
Polygon or Base) for a paid tiercheck_payment — poll until the invoice settles, then receive a
scoped API keyget_signals again, now authenticatedNo dependencies, no API key required to run steps 1–5 (only step 6/7 need a
real payment to complete — the script prints the payment instructions and
exits there unless you pass --pay, which is left as a manual step since it
requires a funded wallet).
https://mcp.alphabotpro.cloud/v1/mcp (JSON-RPC 2.0 over HTTPS, stateless).well-known/mcp/server-card.json.well-known/ai-plugin.jsonopenapi.json · docsMost "AI-ready" APIs still assume a human signs up, enters a card number, and pastes a key into a config file. That breaks the moment the buyer is an autonomous agent. This server lets an agent discover the price, pay in stablecoins, and start pulling data — all inside one continuous session.
Full write-up: https://alphabotpro.cloud/blog/post/how-ai-agents-pay-for-data-with-crypto-x402-mcp-example
Trading signals returned by this API are informational tools, not financial advice, and not a guarantee of profit. Trading carries risk of loss. AlphaBotPro is not affiliated with the Model Context Protocol project or the x402 project — this repo is an independent example of using both.
MIT — see LICENSE.