Bitroad

by bitroadai

Not rated
GitHub

About

Marketplace for AI agents: buy goods and services under spending caps, with returns and disputes.

Details

Author
bitroadai
Categories
Other, AI

Setup

Install Bitroad in your MCP client (Claude Desktop, Cursor, Windsurf, and others).

Repository: https://github.com/bitroadai/bitroad-mcp

Follow the installation instructions in the repository README, then restart your MCP client.

Bitroad is a marketplace built for AI agents. Your agent searches a catalogue of goods and services, places orders under spending caps you set, tracks delivery, and handles returns and disputes, all through the Model Context Protocol.

Endpoint:https://app.bitroad.ai/api/v1/mcp

Transport is spec-compliant Streamable HTTP with JSON-RPC 2.0. Auth is OAuth 2.1 with dynamic client registration and PKCE, so most clients need nothing more than the URL above.

- Website:https://bitroad.ai
- Documentation:
https://bitroad.ai/docs
- Sign up:
https://buy.bitroad.ai/sign-up
- Create a buyer account at
buy.bitroad.ai.
- Add the endpoint to your MCP client and approve the consent screen.
- Your agent can now browse and read orders immediately.
- To let it spend, add a card and set delegation caps in your dashboard. Until you do, there is no purchase path at all.

Spending is bounded by three caps you control: per transaction, per day, and total. A purchase above any cap is refused outright, with a reason ofper_tx_cap_exceeded,daily_cap_exceededortotal_cap_exceeded. Separately, you can set a confirmation threshold: a purchase at or above it is allowed but returnsconfirmation_requiredwith a token, and needs your explicit sign-off before it proceeds. Agents never see card details; a card can only be added by you through Stripe hosted checkout.

There are three shapes. Pick the one that matches your client.

# Claude Code claude mcp add --transport http bitroad https://app.bitroad.ai/api/v1/mcp # Gemini CLI gemini mcp add --transport http bitroad https://app.bitroad.ai/api/v1/mcp

Run the client and trigger the OAuth flow (/mcpin Claude Code, automatic in Gemini CLI), then approve on the Bitroad consent screen.

Cursor, Claude Desktop, Cline, Windsurf, LibreChat and most other MCP clients take a JSON block:

{ "mcpServers": { "bitroad": { "url": "https://app.bitroad.ai/api/v1/mcp" } } }

The client discovers OAuth on first use.

Claude.ai (Settings, then Connectors), ChatGPT (developer mode custom connectors), and Copilot take the endpoint as a pasted URL:
- Open the client's connector settings.
- Add a connector with URLhttps://app.bitroad.ai/api/v1/mcp.
- Approve the Bitroad consent screen when prompted.

For headless clients and your own agent code, mint an agent key at/buyer/instances/newand send it as a header:

curl https://app.bitroad.ai/api/v1/mcp \ -H "Authorization: Bearer br_ik_..." \ -H "Content-Type: application/json" \ -H "Accept: application/json" \ -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

Keys are shown once. Revoke them from the same page.

Calltools/listfor the live catalogue with full JSON Schema.tools/listreturns the whole catalogue to every caller; your account type is enforced when a tool is called, not when it is listed. Buyer and seller are separate account types and one email can only be one of them, so a buyer calling aseller_tool is refused.

Listings, stock, orders, shipping and tracking, returns, and review responses, under theseller_prefix.

A quote-based marketplace for work rather than goods, under theservices_*prefix: request a quote, accept it, and funds are held in escrow until you accept the deliverable.

The catalogue also carriesenvelopes_listandenvelopes_get, a preview surface that is switched off on the hosted service. They appear intools/listbut return a not-found error when called.

Buying a product is a two-step flow.purchase_create_intentreserves stock and snapshots price, VAT and shipping, thenpurchase_confirm_intentcharges and creates the order. Intents expire after 15 minutes. All monetary values are integer pence.

Write tools accept an optional_meta.idempotencyKey. Passing one gives you full replay semantics on retries. If your client cannot set it, the server generates one so the call still succeeds.

{ "jsonrpc": "2.0", "id": 3, "method": "tools/call", "params": { "name": "purchase_create_intent", "arguments": { "product_id": "...", "quantity": 1 }, "_meta": { "idempotencyKey": "intent-abc-123" } } }

This repository holds theserver.jsonrecord published to the official MCP registry under theai.bitroadnamespace.

AI agent commerce marketplace — register, list services, buy and sell capabilities with real payments via Stripe.

Enable AI Agents to purchase anything in a secure way using Fewsats

An MCP server that lets AI agents discover, hire, and pay other AI agents — the exchange where agents are the customer.

Hosted MCP server for Amazon sellers using Claude, ChatGPT, and other AI clients.

AI agent API marketplace-discover, call, and pay for services with USDC payments. List your own AI services and earn money per call.

AI agent commerce: buy gift cards, top up phones, and get travel eSIMs with Bitcoin, Ethereum, Solana, USDC, USDT, Litecoin, Dogecoin, and 15+ cryptos on Base, Polygon, Arbitrum, Tron, and more. No account, no CLI install, no API key — connect via MCP or let your agent pay autonomously with x402.

An MCP server enabling AI agents to pay for services using HTTP 402 requests and MCP tools.

Agentic commerce infrastructure for AI agents. MCP-native product discovery, contextual ad matching, and purchase facilitation with European privacy compliance (nDSG/GDPR).

Living directory of x402 payment endpoints with curated recipes. Agents discover, evaluate, and execute paid workflows via MCP.

MCP server enabling AI agents to autonomously hire human specialists. 18 tools: registration, specialist search by skills/availability/price, profiles, crypto wallet payments, and scheduling.

No reviews yet — be the first

Sign in to leave a review

Use Google, GitHub, or an email account so ratings stay tied to real people.

Email sign in

No reviews posted yet.