Skip to main content
Glama
Hady-Achkar

chain-context-mcp

by Hady-Achkar
README.md
# chain-context-mcp

A lightweight **MCP server** that gives AI agents and Claude Desktop **live crypto market data** —
current price, 24h stats, and top movers — over the Model Context Protocol. Pay-per-call ready (x402),
with a free tier for discovery.

**Live demo server:** `https://chain-context-mcp.hadi-ai.workers.dev`

## Why
Most MCP data servers are read-only wrappers that 403 from cloud IPs or need API keys. This one runs on
Cloudflare Workers, uses a keyless public source, edge-caches responses, and is built to be **paid per
call by autonomous agents** via [x402](https://www.x402.org) — machines paying machines, no signup.

## Tools
| Tool | Args | Returns |
|------|------|---------|
| `get_token_price` | `id` (e.g. `bitcoin`, `ethereum`, or `BTC`) | USD price + 24h % change |
| `get_market_snapshot` | `id` | price, 24h volume, high, low, open, change |
| `get_trending` | — | top major-token movers by 24h % change |

## Connect from Claude Desktop
Add a Streamable-HTTP MCP server pointing at the URL above, then ask:
> "What's the price of solana right now?"

## Quick test
```bash
curl -s -X POST https://chain-context-mcp.hadi-ai.workers.dev \
  -H 'content-type: application/json' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"get_token_price","arguments":{"id":"ethereum"}}}'
```

## Run / deploy your own
```bash
npm install
npm test                 # pure-logic tests (gating)
npm start                # local Node server on :8788
wrangler kv namespace create MCP_USAGE   # paste id into wrangler.toml
wrangler deploy          # deploy to Cloudflare Workers
```

## Monetization
- **Free tier:** `FREE_CALLS_PER_DAY` per client (KV-tracked), consumed only on a successful response.
- **x402:** set `X402_PAYTO` to your USDC wallet; calls beyond the free tier return HTTP 402 with payment
  requirements, and a valid `X-Payment` header is served. (Plug in an x402 facilitator to verify settlement.)

## License
MIT