GTM Health Scorecard v2

Diagnose any company's GTM infrastructure health from public data. Zero dependencies.

Input: A company domain (e.g., ramp.com) Output: 3 red flags → 7-dimension score → diagnosis → priority fixes With :prospect: All of the above + best contact + outreach DM draft

No pip install. No gather.py. Core scorecard needs no API keys — just web search. Two optional Exa boosters (SPA job-description bodies + people discovery) activate only if EXA_API_KEY is set; without it the scorecard runs fully on web search. Note the firewall distinction: Exa /contents returns raw page text (eyeball-able — fine to cite), but Exa /search structured entity fields are provider-asserted — treat them like any provider claim under Step 0.6, never as confirmed stack.

Step 0: Load GTM Context

If .agents/gtm-context.md exists, read it for context about your own company, ICP, and GTM motion. This helps tailor the fixes and outreach to your sales motion. It does NOT feed the diagnosis (see firewall below).

If it doesn't exist, the scorecard still works — it just produces generic fixes instead of ones calibrated to your offering.

Step 0.5: Evidence Firewall (READ THIS — non-negotiable)

The scores, red flags, diagnosis, and evidence section must be derived ONLY from public data gathered during THIS run (web search, the company's site, live JDs). Nothing else.