Skip to main content
Query blockchain data across 20+ chains directly from your AI assistant.

​
How It Works

When you ask questions about blockchain data, Claude uses these MCP tools:
  1. find_tables - Discovers relevant tables for your question
  2. get_table_schema - Gets column details to write accurate SQL
  3. run_query - Executes the SQL and returns results
  4. query_results_with_duckdb - Analyzes large results locally

​
Example Conversations

​
Basic Query

You: What were the top 10 wallets by ETH transfer volume yesterday? Claude: I’ll query the Ethereum transfer data… [Claude finds tables, writes SQL, executes query, returns results]

​
Complex Analysis

You: Compare DEX trading volume across Ethereum, Arbitrum, and Base for the last 7 days Claude: I’ll query each chain’s DEX data and compare them… [Claude runs multiple queries and synthesizes results]

​
Follow-up Analysis

You: Can you break that down by protocol? Claude: Let me add protocol grouping to the analysis… [Claude modifies the query based on previous results]

​
Available Tools

​
run_query

Execute SQL and wait for results. Returns metadata plus preview rows. 
Tool: run_query
Parameters:
  - sql: The SQL query to execute
  - name: Human-readable name for the query
  - previewRows: Number of preview rows (0-100, default 5)
  - timeoutMs: Max wait time (default 120s, max 300s)

​
execute_sql

Execute SQL and save as a named query for future reference. 
Tool: execute_sql
Parameters:
  - sql: The SQL query to execute
  - name: Name for the saved query
  - description: Description of what the query does

​
get_query_results

Fetch results from a completed query run. 
Tool: get_query_results
Parameters:
  - queryRunId: ID of the query run
  - format: Output format (json or csv)
  - limit: Max rows to return
  - offset: Rows to skip

​
query_results_with_duckdb

Analyze query results locally using DuckDB SQL. Useful for aggregations and transformations without fetching all data. 
Tool: query_results_with_duckdb
Parameters:
  - queryRunId: ID of the query run to analyze
  - sql: DuckDB SQL query (use 'results' as table name)
  - maxRows: Max rows to return (default 1000)

​
run_sql_for_download

Execute SQL and get a download URL for the results. 
Tool: run_sql_for_download
Parameters:
  - query: The SQL query to execute
  - format: Export format (csv or json)

​
Query Management

​
list_queries

List all saved queries available for execution.

​
list_query_executions

List recent query runs with names, descriptions, and status. Useful for reusing previous queries.

​
get_query

Get details of a specific saved query including its SQL.

​
Tips for Better Results

​
Be Specific

Instead of: “Show me Ethereum data” Try: “What were the top 10 NFT collections by sales volume on Ethereum yesterday?”

​
Specify Time Ranges

Instead of: “Show me DEX trades” Try: “Show me Uniswap v3 trades on Ethereum from the last 24 hours”

​
Ask for Explanations

“Query the top wallets by transaction count, and explain what each column means”

​
Request Formats

“Query gas prices for the last week and format the results as a markdown table”

​
Supported Chains

Flipside provides data for 20+ blockchains including:
  • Ethereum, Arbitrum, Base, Optimism, Polygon
  • Solana, Bitcoin, Avalanche, BSC
  • Cosmos, Osmosis, Sei, Injective
  • NEAR, Aptos, Sui
  • And many more…
Use find_tables to discover what data is available for each chain.