MCP App Store
by KeenEthics
General Index icon

General Index MCP Server for Claude and ChatGPT

By General Index

Overview

Ask an AI assistant for a commodity price today and you get something scraped, stale or invented — a number you can’t trust. This connector puts General Index’s published benchmarks into the conversation instead: over 6,200 independent, assessed price indexes across crude, refined products, natural gas, LPG/NGL, marine fuels, carbon, biofuels, SAF, hydrogen and tank storage, built from data contributed by more than 200 market participants. Connect once and your assistant can find the right benchmark from a plain-English question and pull the live number — no integration project, no code, no code lists to memorise. Ask for “Singapore 10ppm gasoil”, “European carbon” or “US Gulf Coast jet fuel” and get the matching indexes with their details, then the latest value, the history, or the full curve. Read-only and published data only. You see the indexes you’re entitled to. Requires a General Index subscription with MCP access enabled.

How to connect

  1. Claude: Open the General Index listing in the Claude connectors directory and choose Connect. Claude listing General Index exposes 3 tools here.
  2. ChatGPT: Open the General Index listing in the ChatGPT app directory and choose Connect. ChatGPT listing General Index exposes 3 tools here.

Tools

query_index_db

ChatGPT
Execute natural language queries against the GX index pricing time-series database. This tool is for questions about PRICE DATA — actual index values, datapoints, time-series analysis, and cross-product price analytics. It queries prod_indexes.data (time-series price rows) and prod_indexes.meta (product metadata joined at query time). Use this tool for: - Price lookups and time-series: "What were WTI prices last week?" - Aggregations over price data: "Monthly averages for GX0013984 in 2025" - Cross-product price analytics: "Which crude grade moved most this week?" - Datapoint counts in the price DB: "How many index datapoints were published in the past 2 years?" - Volatility / trend analysis: "Analyze European gasoil volatility in Q4 2025" Do NOT use this tool for catalog questions about which products or codes exist — those belong to search_catalog (e.g. "how many codes are in the Compliance Carbon module?", "list all forward curve products in Europe"). IMPORTANT: Queries are restricted to prod_indexes.data and prod_indexes.meta tables.

request_human_assistance

ChatGPT
Request human assistance when you cannot help the user directly. Files a support request with the relevant General Index team. The request is delivered as an email and a Teams notification; someone from the named team follows up with the user directly. Use this tool when: - The user explicitly asks to speak with a human or a team member. - You cannot find relevant data after multiple search attempts. - The user needs account support, billing help, or administrative assistance. - The user reports a bug or technical issue with the platform. - The user provides feedback that requires team review.

search_catalog

ChatGPT
Search and query the GX product catalog — the authoritative source of which GX codes exist, what modules/commodities/hubs they belong to, and product-level metadata. Use mode="advanced" for any catalog question that requires reasoning over the full product set: - LISTINGS: "Show me all products we offer for Natural Gas and LNG" "List all forward curve products in Europe" "What NGL products are available?" - COUNTS / COVERAGE: "How many codes are in the Compliance Carbon module?" "How many crude products do we have in North America?" "How many forward curve products exist for European gas?" Advanced mode runs a pandas agent over the live catalog DataFrame and returns either the matching product set (listing) or a plain-text count answer (analytical). num_results is IGNORED in advanced mode — there is no cap on results. Use mode="quick" (default) for fast product lookups by name — e.g. resolving "WTI" or "Henry Hub" to a product record before querying index data. num_results controls how many vector-search results to return (default 5, max 50). Set id_search=True when the query is one or more exact GX codes (e.g. "GX0011275" or "GX0018107, GX0000971, ..."). This resolves codes directly from the vector DB (keyed by code) with no semantic search or reranking — far faster and exact. Semantic search cannot match a bare code, so always prefer id_search for code lookups. num_results is ignored (every requested code is returned). Each result carries other_fields.symbols — the product's assessment series as a list of {period_type, time_ref, period_min, period_max} tuples, i.e. the REAL co-occurring (PeriodType, TimeRef) combinations for that code (never a lossy cross-product; a code that trades Month/1630 + Prompt/1730 will not imply a phantom Month/1730). Use these to pick the exact PeriodType/TimeRef to pass to query_index_db's symbols argument. In advanced mode, add "symbols" to a result set's "fields" list to include it. This tool is about the CATALOG (what products exist). For questions about actual index prices, time-series data, or datapoints in the index database, use query_index_db instead.

query_index_db

Claude

request_human_assistance

Claude

search_catalog

Claude

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/general-index"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/general-index)