Web3 Chain Radar: Install, Source and Security | FunnelSlayer

Web3 Chain Radar

Published by 985-web3 in web3

Audit unavailable1 installs

What this skill does

Use the web3-chain-radar MCP server to scan on-chain tokens, classify narratives, inspect momentum, liquidity, and basic safety risks. 当用户想查链上新币、GMGN 排行、叙事分类、动量信号或基础安全风险时使用。 Requires the `web3-chain-radar` MCP server to be configured first.

Add Web3 Chain Radar to your agent

Review the source and files first. When you are ready, copy the prompt instruction or use the CLI command supported by your environment.

Install with a prompt

Paste this into a compatible coding agent:

add this skill "web3-chain-radar" from https://github.com/985-web3/web3

Install with the CLI

Run this command in a controlled environment after reviewing the repository:

npx skills add https://github.com/985-web3/web3 --skill web3-chain-radar

Skill instructions

web3-chain-radar

What This Skill Is

This is a thin Hermes-style skill wrapper for the web3-chain-radar MCP server.

It is meant to be dropped into a skills directory so the assistant knows:

  • when to use the radar
  • which MCP tools to call
  • how to present the results
  • what not to do

This skill does not replace the MCP server. It depends on the web3-chain-radar MCP service being installed and connected.

Prerequisite

Before using this skill, make sure the MCP server is available as web3-chain-radar.

If the MCP server is missing, explain that this skill needs the web3-chain-radar-mcp service installed and configured first.

When To Use

Use this skill when the user wants to:

  • scan new or active tokens on ETH, BSC, Base, or Solana
  • inspect GMGN-ranked tokens
  • classify narratives such as Musk/Trump, Binance/CZ, celebrity or viral themes, or emerging themes
  • check momentum, liquidity, buys and sells, and basic safety signals
  • get ranked research candidates instead of direct trading execution

Tools

scan_onchain_narratives

Use this as the primary tool.

Recommended default call:

scan_onchain_narratives(
  chains="eth,bsc,base,sol",
  include_safety=true,
  strict_safety=false,
  include_profiles=false,
  max_results=20
)

Adjust parameters only when the user narrows scope.

classify_token_narrative

Use this when the user asks about a token name, symbol, or theme classification without needing a full market scan.

Output Guidance

When summarizing results, prioritize:

  • token name, chain, and address
  • narrative category and star level
  • market cap, liquidity, and short-term price change
  • safety result and whether the check succeeded
  • why the token ranked highly

Keep the output concise and structured for decision support.

Boundaries

This skill is for scanning and analysis only.

  • Do not trade
  • Do not connect wallets
  • Do not handle private keys
  • Do not present results as guaranteed profit

If the user asks to buy, sell, or place trades, treat the radar output as research context only.

Files included

  • SKILL.md

More skills