GRID

What is GRID?

GRID is the spreadsheet engine purpose-built for agentic products. Instead of burning tokens and generating brittle code every time an LLM needs to read or modify a spreadsheet, GRID gives your agent a structured interface to call spreadsheet operations directly. It runs deterministic calculations that match Excel and Google Sheets, provides ready-made UI components for human interaction, and integrates quickly via npm. Whether you're building a financial analyst tool, an AI-powered spreadsheet editor, or a custom Excel add-in, GRID lets your agent handle complex workbook logic with a single function call—fast, accurate, and sandboxed.

What are the features of GRID?

• Deterministic Spreadsheet Engine: Headless calculation engine that matches Excel and Google Sheets output exactly, with no code generation needed from the LLM.
• LLM-Optimized Tool Interface: Agents call readCell, writeCell, goalSeek, and other structured operations instead of writing openpyxl or pandas code.
• Ready-Made UI Components: Includes a themeable Spreadsheet Viewer and Editor that feels native inside your product, plus an Excel add-in capability.
• Sandboxed Execution: Catches errors before they touch the real file—critical for use cases where corrupting a workbook is unacceptable (e.g., investment banks).
• Fast Integration: Install from npm with a free evaluation license, add GRID's skills for agents via npx skills add, and go to production quickly with commercial licensing.
• Fully Compatible with .xlsx: Drag-and-drop Excel files, edit cells, recalculate hundreds of cells in real time, and download updated workbooks.

What are the use cases of GRID?

• Agentic Financial Modeling: An LLM-based analyst uses GRID's goalSeek to find the churn rate needed to hit a $250k ARR target—returns an answer in 1.5 seconds using 86 tokens instead of 11 seconds and 3,240 tokens with generated Python code.
• AI-Powered Spreadsheet Editor: Build a React-based editor where users can upload .xlsx files, edit cells, and download updated versions, with all heavy lifting handled by GRID's engine.
• Custom Excel Add-in: Create an add-in for major investment banks where the model runs domain-specific skills through GRID's sandboxed engine, preventing any corruption of sensitive financial models.
• LLM Platform Spreadsheet Feature: Integrate GRID as the core spreadsheet engine for a leading LLM provider, enabling reliable read/write/recalculate workflows at scale without code generation.
• Real-Time Interactive Models: Display interactive sliders that recalculate hundreds of cells in a property-value model instantly, with all logic running deterministically in the browser or server.

How to use GRID?

• Install the Engine: Run npm install @grid-is/spreadsheet-engine in your project. This gives you the headless engine with a free evaluation license.
• Add Agent Skills: For LLM agents, execute npx skills add GRID-is/skills -g --all to install GRID's structured tool calls for reading, writing, and recalculating spreadsheets.
• Integrate UI Components: Use the Spreadsheet Editor or Viewer React components to display and edit spreadsheets inside your app. Configure them to use the engine for all heavy operations.
• Connect Your Agent: Have your LLM call engine methods like engine.write("A2", 125000) or engine.goalSeek({...}) instead of generating Python or JavaScript libraries. The engine handles formula recalculation deterministically.
• Test and License: Experiment with the free evaluation license. When you're ready to ship, contact GRID for commercial licensing and production support.
• Handle File I/O: Your UI can accept .xlsx uploads (drag-and-drop), load them into the engine, let users or agents edit cells, and export the updated workbook as a downloadable file.