HubSpot Deal Health Scorer

What it does

HubSpot Deal Health Scorer scans a HubSpot pipeline and computes a transparent 0-100 health score for every open deal (activity age 50 pts, close-date proximity 30 pts, amount 20 pts), classifying each as healthy, stale, at_risk or critical. For every problem deal it generates a concise AI-written follow-up suggestion — the LLM narrates only, every score is code-derived. Read-only: never writes back to HubSpot deals. Optionally posts a formatted health summary to a Slack channel. Ideal for sales managers and RevOps teams who want an automated, explainable early-warning system for their pipeline instead of manually scanning deal lists.

Example prompts

• Run the full deal health scan for my HubSpot pipeline
• Show me only the at-risk and critical deals with suggested actions
• Which deals are stale with no recent activity?
• Score these deals and classify them by risk

Tools (7)

Tools the agent exposes — your AI client calls them automatically when it needs them.

• run_full — Fetch all open HubSpot deals, compute health scores, generate AI follow-up suggestions, and return the full deal health report as JSON.
• score_deals — Given a list of raw HubSpot deals, run deterministic health scoring (no LLM) and return scored deals with risk classification.
• get_at_risk_deals — Run the pipeline health check and return only at-risk and critical deals with suggested actions.
• get_stale_deals — Run the pipeline health check and return only stale deals (no recent activity).
• list_capabilities — List available tools, scoring algorithm details, credential requirements and thresholds.
• plan_inputs — Plan inputs for any tool: returns questions, schema and a ready-to-edit example.
• discover_intent — Intent-discovery front door: describe your goal in natural language and get a concrete proposed input.