Agent-Rankingscore
Up from a grade D on the first scan eight weeks ago.
Score by category, before vs. after
Current issue breakdown
Score history
What the agent found this week.
Flagged automatically during the scheduled re-scan, ranked by how much it affects agent usability.
The OAuth flow for API access still requires a manual approval email - no self-serve sandbox key exists yet.
Pricing page content changed last week; the generated pricing.json has not been regenerated to match.
Two MCP tool descriptions are shorter than recommended, which may reduce how confidently an agent invokes them.
A new endpoint, /v1/reports/export, was detected in the OpenAPI spec and is not yet wrapped as an MCP tool.
Before and after, by category.
| Category | Before | After | Remediation applied | Status | |
|---|---|---|---|---|---|
| Structured data | 31 | → | 92 | Generated schema.org markup for pricing and product pages | Resolved |
| Agent manifest | 0 | → | 100 | Published agents.json and llms.txt at site root | Resolved |
| API accessibility | 44 | → | 88 | Wrapped 12 REST endpoints as MCP tools | Resolved |
| Auth simplicity | 58 | → | 74 | Added sandbox key issuance flow | In progress |
| Pricing clarity | 62 | → | 95 | Published machine-readable pricing.json | Resolved |
Every site we've evaluated, public.
A running log of businesses that have gone through an evaluation, color-coded by whether they currently pass - full list on its own page since it grows fast.
Want a report like this for your own site?
Submit your URL and get your first readiness score.