web3-researcher

Web3-native research routing for AI agents — finds the right data source automatically instead of relying on general search engines.

An Agent Skill that routes any Web3 research query to the most accurate native source: DeFiLlama for DeFi data, CoinGecko for prices, Dune for onchain metrics, chain explorers for contracts, Farcaster for community sentiment, and more.

---

What it does

1. Classifies the query — identifies what type of Web3 research is needed
2. Routes to the right source — picks the best Web3-native data source, not Google
3. Combines sources — for complex queries that need multiple angles
4. Outputs a structured report — findings, sources used, and confidence level

---

Sources covered

---

Install

https://github.com/vivu16/multi-search-engine/blob/main/SKILL.md

Or fetch raw:

curl -s https://github.com/vivu16/multi-search-engine/blob/main/SKILL.md

---

Quick start

Load the skill then ask:

“Research how Aerodrome is performing on Base right now.”

“What are the best USDC yields on Base?”

“Is this contract safe? 0x1234…abcd on Base”

“What are Web3 builders saying about Superfluid?”

---

File structure

web3-researcher/
├── SKILL.md                           # Core skill — load this into your agent
├── README.md                          # You are here
├── QUICKSTART.md                      # Get running in 5 minutes
├── CHANGELOG.md                       # Version history
└── references/
    ├── SOURCE_DIRECTORY.md            # Full directory of all Web3 sources
    ├── QUERY_ROUTING.md               # Decision trees for complex queries
    └── OPERATORS.md                   # Advanced search operators per source

---

Built for the ProductClank ecosystem

Registered with the ProductClank Skill Registry and follows the Agent Skills standard (v1, Anthropic).

Useful links:

• ProductClank Agents
• Skill Registry — Season 6
• Official ProductClank agent skills

---

Author

ELARA VOSS

---

License

MIT