sp-global

S&P Global Plugin

This plugin brings S&P Global's financial data and analytics directly into your AI workflow via a set of pre-built skills. It is designed for financial professionals who want AI-assisted research, analysis, and document generation grounded in authoritative S&P Global data.

The skills are built on open standards (MCP) and are designed to work across AI platforms and agent frameworks. While the plugin follows the Claude Cowork standard, all skills and the underlying data layer are platform-agnostic. If you want to use these skills in another environment, please do.

We recognize that every organization has unique needs. These skills are starting points to get you started on your finished products. We encourage you to adapt the prompts, outputs, and workflows to fit your firm's specific processes, templates, and data needs.

The skills in this plugin are provided as-is. Generated outputs are not guaranteed to be correct. Always verify outputs generated by an LLM for correctness.

Skills Contained

Tearsheets

Requires: S&P Global LLM-ready API subscription

Generates a formatted, one-to-two page company tearsheet as a Word document, populated with live data from S&P Capital IQ. Supports four audience types, each optimized for a different use case:

• Equity Research: Investment thesis snapshot for buy-side/sell-side analysts
• Investment Banking / M&A: Company profile in transaction context
• Corporate Development: Acquisition target profile for internal strategic teams
• Sales / Business Development: Client meeting prep for commercial teams

Example prompt: "Generate a business development tearsheet for Palantir."

Funding digests

Requires: S&P Global LLM-ready API subscription

Summarizes funding activity within a sector or for a specific company, drawing on S&P Capital IQ transaction data. Useful for market mapping, pitch preparation, and competitive intelligence.

Example prompt: "Summarize recent transactions in the data infrastructure space”

Earnings Previews

Requires: S&P Global LLM-ready API subscription

Generates a structured earnings preview for an upcoming report, including consensus estimates, recent guidance, analyst sentiment, and key things to watch — all sourced from S&P Capital IQ.

Example prompt: "Give me an earnings preview for Salesforce."

How to Use

The plugin and skills require access to S&P Global data to work with, either Capital IQ Pro or S&P Global LLM-ready API subscriptions.

The LLM-ready API is easy to integrate in Claude or other applications via its MCP server. Follow these steps to set it up.

In Cowork

You'll need a paid Claude plan (Pro, Max, Team, or Enterprise) and the Claude Desktop app for macOS or Windows.

1. Open Claude Desktop and navigate to the Cowork tab
2. Click Customize with Plugins
3. In Browse Plugins, select Personal
4. Click the plus sign “+” to add plug in
5. Authenticate with your S&P Global credentials when prompted

Once installed, skills activate automatically when relevant – just describe what you need in natural language. You can also invoke specific skills explicitly by typing / in the chat to see available commands.

To customize the plugin for your firm's workflows, templates, or terminology, click Customize while viewing the installed plugin. We encourage this; the defaults are a starting point, not a prescription.

In Claude Desktop (individual skills)

To install individual skills in Claude Desktop without the full plugin:

1. Open Settings
2. Navigate to Capabilities → Skills
3. Click Add
4. Upload the skill file(s) from this repository

Skills are available immediately in your Claude Desktop sessions once uploaded. You can install one or several depending on your needs.

In Claude Code (individual skills)

Follow the instructions on the Claude Code documentation.

Other Platforms

The skills in this repository are markdown files. Any AI platform that supports custom instructions, system prompts, or knowledge file uploads can use them – the mechanism varies by platform but the principle is the same: load the skill content so the model has it as persistent context.