An MCP (Model Context Protocol) server that wraps the AdValorem MEV Intelligence API, giving AI agents real-time access to DeFi MEV data: liquidation wave predictions, searcher leaderboards, builder routing recommendations, and a risk-enriched opportunity feed.
Run directly with npx (no install required):
npx @advalorem/mev-intelligence-mcpOr install globally:
npm install -g @advalorem/mev-intelligence-mcp
mev-intelligence-mcpAdd this server to your MCP client config (e.g. Claude Desktop claude_desktop_config.json):
{
"mcpServers": {
"mev-intelligence": {
"command": "npx",
"args": ["@advalorem/mev-intelligence-mcp"]
}
}
}| Tool | Description | Endpoint |
|---|---|---|
mev_feed |
Real-time risk-enriched MEV opportunity feed across Aave V3, Spark, Morpho (L1), and Aave V3 (Base) | GET /preview/feed |
mev_liquidation_waves |
Predicted liquidation waves — borrowers about to become liquidatable based on oracle pressure | GET /preview/liquidation-waves |
mev_searcher_leaderboard |
Top liquidator searchers ranked by landed fires, attempts, and land-rate over the last 4 hours | GET /preview/searcher-leaderboard |
mev_builder_recommendation |
Best builder to route a liquidation bundle through, based on acceptance rate and volume | GET /preview/builder-recommendation |
mev_feed
| Parameter | Type | Default | Description |
|---|---|---|---|
limit |
number | 50 |
Max items to return |
mev_liquidation_waves
| Parameter | Type | Default | Description |
|---|---|---|---|
window_min |
number | 30 |
Lookback window for oracle pressure (minutes) |
top_n |
number | 10 |
Number of waves to return |
mev_searcher_leaderboard
| Parameter | Type | Default | Description |
|---|---|---|---|
window_min |
number | 240 |
Lookback window (minutes, 4h default) |
top_n |
number | 25 |
Number of searchers to return |
mev_builder_recommendation
| Parameter | Type | Default | Description |
|---|---|---|---|
window_min |
number | 60 |
Lookback window for builder stats (minutes) |
All MCP tools use the free preview endpoints (/preview/*):
- Rate limited to 60 requests/hr per IP
- Responses are cached for ~5 minutes
- Returns truncated payloads
For full data, no rate limits, and lower latency, upgrade to the paid intelligence tier via x402 micropayments ($0.10–$0.50 USDC on Base per call):
| Paid Endpoint | Price |
|---|---|
GET /intelligence/feed |
$0.10 USDC |
GET /intelligence/liquidation-waves |
$0.50 USDC |
GET /intelligence/searcher-leaderboard |
$0.25 USDC |
GET /intelligence/builder-recommendation |
$0.25 USDC |
See the full catalog and payment instructions at: https://mev.advalorem.io/intelligence/catalog
Built by AdValorem. Data is derived from internal on-chain pricer and dispatcher state — not resold third-party feeds.
- API documentation: https://mev.advalorem.io/intelligence/openapi
- Intelligence catalog: https://mev.advalorem.io/intelligence/catalog
- Contact: val@advalorem.io
- GitHub: https://github.com/ValoremFoundation/mev-intelligence-mcp
MIT