What Rug Radar Does

Rug Radar is a Solana-focused MCP server for fast token and wallet due diligence. Most users should use the already deployed remote MCP endpoint at https://mcp.rugrdr.xyz/mcp inside Claude or any MCP-compatible client. Self-hosting is supported, but it is the advanced path, not the default path.

It is designed to answer a narrow but high-value set of questions quickly:

Is this token risky?
Which specific failure modes are visible right now?
Does this wallet look like a sniper, bot, whale, insider, or normal retail flow?
Which of two tokens looks healthier on a risk-adjusted basis?
What changed since the last snapshot?

Who It Is For

Retail traders who need a first-pass risk read before entering or sizing a trade
Researchers who want structured snapshots instead of manually stitching together APIs
Communities and moderators who need short, repeatable token reviews
Agent builders who want a remote MCP endpoint with deterministic tool outputs

Default Way To Use Rug Radar

For most users, the workflow is:

connect your MCP client to https://mcp.rugrdr.xyz/mcp
let the client discover the Rug Radar tools
use plain-language prompts like Analyze BONK or Compare BONK vs WIF

You do not need to clone the repository or run the server locally just to use Rug Radar.

What The Project Includes

A TypeScript MCP server in src/
Six MCP tools for token analysis, wallet analysis, comparison, and monitoring
A Docusaurus docs app in apps/docs/
Optional Postgres-backed watchlists and snapshots
Optional Upstash Redis alert deduplication
Public HTTP metadata endpoints at /, /info, and /health

Two Ways To Use The Project

1. Hosted MCP

Best for almost everyone.

no local server setup
no API keys
connect once and start using tools

2. Self-hosted MCP

Best only if you need:

your own infrastructure
custom auth or private deployment controls
code changes or integration work
your own database-backed monitoring environment

MCP Tools

analyze_token: Full token due diligence report by address or token name
scan_risk: Fast red-flag scan with triggered and passed checks
analyze_wallet: Wallet behavior classification based on activity heuristics
compare_tokens: Side-by-side comparison across key token quality metrics
watch_token: Persist monitoring thresholds and baseline snapshots
get_token_changes: Compare the current snapshot against the previous one and emit alerts

What Rug Radar Optimizes For

Speed over exhaustive chain indexing
Structured output over long-form prose
Partial answers instead of complete failure when one upstream source is down
Deterministic scoring so outputs are easier to reason about