HTTPayer MCP
Enables AI agents to call x402-gated APIs using a central credit balance, abstracting away blockchain complexity and payment proofs. It provides tools to fetch data from payment-required endpoints, check usage balances, and simulate transaction costs.
README
@httpayer/mcp
MCP (Model Context Protocol) server for HTTPayer. Lets AI agents call x402-enabled APIs using credit balance — no wallets, no blockchain, no Web3 knowledge required.
- Dashboard & API keys: app.httpayer.com
- npm: @httpayer/mcp
- GitHub: httpayer/mcp
Quickstart
With an AI agent (recommended)
Paste this into any MCP-compatible agent (Claude Code, Cursor, Windsurf, OpenCode...):
Set up https://httpayer.com/skill.md
The agent detects your environment and handles everything automatically.
Without an agent (manual)
1. Run setup:
npx @httpayer/mcp setup
Get your API key at app.httpayer.com when prompted.
Flags:
| Flag | Description |
|---|---|
--key sk-live-... |
Provide key non-interactively |
--client <name> |
Target client: claude-code, claude-desktop, cursor, windsurf, opencode, zed, cline, warp, codex |
--scope user|project |
Claude Code scope (default: user) |
--yes / -y |
Skip all prompts |
--update-key |
Replace existing key |
2. Add to your client:
Claude Code:
claude mcp add httpayer --scope user -- npx -y @httpayer/mcp@latest
Claude Desktop (~/Library/Application Support/Claude/claude_desktop_config.json on macOS, %APPDATA%\Claude\claude_desktop_config.json on Windows):
{
"mcpServers": {
"httpayer": {
"command": "npx",
"args": ["-y", "@httpayer/mcp@latest"]
}
}
}
Cursor (.cursor/mcp.json), Windsurf (.windsurf/mcp.json), Cline (.cline/mcp_settings.json):
{
"mcpServers": {
"httpayer": {
"command": "npx",
"args": ["-y", "@httpayer/mcp@latest"]
}
}
}
OpenCode (opencode.json or ~/.config/opencode/config.json):
{
"mcp": {
"httpayer": {
"type": "local",
"command": ["npx", "-y", "@httpayer/mcp@latest"],
"enabled": true
}
}
}
Zed:
{
"context_servers": {
"httpayer": {
"command": {
"path": "npx",
"args": ["-y", "@httpayer/mcp@latest"]
}
}
}
}
3. Restart your client and verify:
Ask your agent: "fetch https://api.httpayer.com/demo/v1/base-weather"
A weather response means HTTPayer is working.
How it works
User prompt
│
▼
AI agent (Claude Code, Cursor, Windsurf...)
│ uses MCP tools
▼
@httpayer/mcp (local MCP server via npx)
│ REST calls with x-api-key header
▼
api.httpayer.com
│ proxy handles x402 payment to target
▼
Target x402-gated API
Runtime flow
- Your client launches the MCP server via
npx -y @httpayer/mcp@lateston startup (stdio transport). - The server reads the API key from
~/.httpayer/mcp-config.json. - The agent receives the tool list and system instructions in its context.
- When the agent calls
fetch, the MCP server forwards the request toPOST https://api.httpayer.com/proxy. - HTTPayer's proxy detects a 402, pays using your credits, retries, and returns the final response.
- The result (status, body, headers) comes back to the agent.
MCP tools reference
get_balance
Check credit balance and daily usage.
Input: none
Example response:
{
"account_id": "account_123",
"mainnet": {
"credits_balance": 50000,
"daily_limit": 100000,
"daily_spend": 15500,
"daily_remaining": 84500
}
}
fetch
Make an HTTP request to any x402-enabled endpoint. Payment is handled automatically.
Input:
| Field | Type | Required | Description |
|---|---|---|---|
url |
string | yes | Target URL |
method |
string | no | GET, POST, PUT, DELETE, PATCH — default GET |
body |
object | no | JSON request body |
params |
object | no | Query string parameters |
headers |
object | no | Additional request headers |
timeout |
number | no | Timeout in seconds, max 120 |
Example response:
{
"status": 200,
"body": { "data": "..." },
"headers": { "content-type": "application/json" }
}
On 502, the response includes webhook_id for async polling.
simulate
Dry-run a fetch. Returns cost estimate without spending credits.
Input: Same as fetch (except timeout).
Example response:
{
"requiresPayment": true,
"proxyFeeBreakdown": {
"targetAmount": 0.01,
"proxyFee": 0.0003,
"totalCreditsCharged": 10.3
}
}
get_topup_link
Returns the dashboard URL to add credits. Show to user when balance is low.
Input: none
check_limits
Check global HTTPayer system daily limits and remaining capacity.
Input: none
get_webhook_status
Poll the status of an async operation. Use when fetch returns a 502 with webhook_id.
Input: webhook_id (string, required)
Status values: pending, success, success_refunded, payment_failed, upstream_error, internal_error, rate_limited
HTTPayer API reference
Authentication: x-api-key: sk-live-... header on all requests.
| Method | Path | Tool |
|---|---|---|
GET |
/v1/credits/balance |
get_balance |
POST |
/proxy |
fetch |
POST |
/proxy/sim |
simulate |
GET |
/limits |
check_limits |
GET |
/webhooks/{id} |
get_webhook_status |
Proxy endpoint
POST https://api.httpayer.com/proxy
{
"api_url": "https://target.example.com/endpoint",
"method": "GET",
"json": { "key": "value" },
"params": { "query": "param" },
"headers": { "Custom-Header": "value" },
"timeout": 30
}
Only api_url and method are required.
Status codes:
| Code | Meaning |
|---|---|
200 |
Success |
402 |
Insufficient credits |
429 |
Rate limited |
500 |
Proxy error |
502 |
Target refused payment — includes webhook_id |
Configuration
API key stored at: ~/.httpayer/mcp-config.json
{ "apiKey": "sk-live-..." }
To update: npx @httpayer/mcp setup --update-key
x402 protocol overview
x402 is an HTTP-native micropayment protocol using the 402 Payment Required status code.
Without HTTPayer:
- Client hits endpoint → gets
402+ payment requirements - Client pays on-chain (requires wallet + USDC)
- Client retries with payment proof
With HTTPayer:
- Client calls
POST /proxy { api_url, method, ... } - HTTPayer detects
402, pays using your credits - HTTPayer retries and returns the final response
All blockchain interaction happens on HTTPayer's side.
Credit system
| Unit | Value |
|---|---|
| 1 credit | 0.001 USDC |
| 1 USDC | 1,000 credits |
| Proxy fee | 3% of target payment |
Top up at app.httpayer.com. Below 100 credits, the agent will prompt you to top up.
Error handling
Setup errors
| Situation | Behavior |
|---|---|
| Key format invalid | Print error, exit 1 |
| Key rejected (401) | Print "API key rejected", exit 1 |
| Network unreachable | Print reason, exit 1 |
MCP tool errors
All errors return isError: true — the server stays alive and the agent gets a readable message.
| Situation | Message |
|---|---|
| No config | "No HTTPayer API key configured. Run: npx @httpayer/mcp setup" |
| API non-2xx | "HTTPayer {status}: {body}" |
| Unknown tool | "Unknown tool: {name}" |
© 2026 HTTPayer Inc.
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.
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.
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.
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.