Scorecard methodology
The scanner loads your website the same way an AI agent would and discovers every WebMCP tool it exposes. It then evaluates the implementation across two dimensions:
The overall grade is a weighted blend of three scoring parameters.
Final grade weighting
Hover a parameter to see how it's scored
An agent reviews your tool surface, judging whether the tools are useful, clearly named, well described, and easy to call. Scored on a strict 1–5 scale.
How much of your site is exposed to agents. Great implementations enable complete user journeys, not just the homepage.
The mechanical hygiene of each tool — the things that make a tool callable without guesswork. Averaged across every tool we find.
snake_case nameEach score maps to a letter grade.
| Grade | Tier | Stars | What it means |
|---|---|---|---|
| A+ | Exceptional | ★★★★★ | Best-in-class agent surface |
| A | Excellent | ★★★★★ | Strong, minor gaps only |
| A− | Very good | ★★★★☆ | Solid surface, a few fixes away |
| B+ | Good | ★★★★☆ | Usable, real room to improve |
| B | Solid | ★★★☆☆ | Works, but gaps an agent will feel |
| B− | Needs work | ★★★☆☆ | Thin or loosely specified |
| C | Early | ★★☆☆☆ | WebMCP detected — the starting line |
A blank grade means one of these results:
document.modelContext, navigator.modelContext, and declarative tool elements.Every tool is sorted into one of three categories — a trust ladder for how freely an agent can call it. The category is assigned automatically by classifying each tool's name, description, and input schema during the scan, so it's inferred, not declared by the site.