Scoring Overview
Every scan produces two scores and a stance classification. The two scores are independent - a site can score high on both. For example, a site may block most AI agents via a well-structured robots.txt (high blocking score) while also having excellent structured data for the agents it does allow (high open score). The UI lets the user toggle between modes.
Outputs
| Output | Range | Meaning |
|---|---|---|
| Open score | 0–100 | How well the site serves AI crawlers |
| Blocking score | 0–100 | How effectively the site blocks AI crawlers |
| Stance | OPEN / MIXED / BLOCKING / UNKNOWN | Derived from robots.txt agent rules |
Check statuses
Each individual check resolves to one of five statuses:
| Status | Meaning | Counts toward score? |
|---|---|---|
pass | Criterion fully met | Yes - full point |
warn | Partially met or best-practice gap | Yes - half point |
fail | Criterion not met | Yes - zero points |
info | Informational only; absence is not a problem | No |
skip | Could not be evaluated | No |
Stance classification
Stance is derived from the ratio of blocked to total known AI agents found in robots.txt:
| Stance | Condition |
|---|---|
BLOCKING | ≥ 75% of known agents are blocked |
OPEN | ≤ 25% of known agents are blocked |
MIXED | Between 25% and 75% blocked |
UNKNOWN | No agent-specific rules found at all |