In large organizations with 200+ variables, the AI Agent and MCP now has a way to know where to focus. You can mark any variable as AI Priority, and Abacum's AI features will treat them as high-priority context when generating answers and analysis.

✨ What's new

Every model is different, and not all variables carry equal weight. AI Priority gives you a direct way to guide the AI Agent toward the metrics that matter most in your model, whether that's ARR, Gross Margin, headcount, or any variable that drives your analysis.

When a variable is marked as AI Priority, the AI Agent and MCP will favor it when selecting context to build a response. In practice, this means the AI is more likely to surface the right numbers, reference the right metrics, and produce more relevant outputs for your specific model structure, without you having to prompt it explicitly every time.

⚙️ How it works

Open any variable in the Model Builder, go to the variable settings in the right sidebar, and toggle AI Priority on. The flag applies to the variable globally, not per version, so it works consistently across your Actuals and Forecast contexts. There is no limit on how many variables you can mark, and only users with edit model permissions can toggle the setting. If no variables in a model are flagged, the AI Agent falls back to its default behavior and considers all variables equally.