How the OpenForBots AI Discovery Audit Works
What the audit checks, how evidence is labelled, how crawler policy is interpreted, and what the results cannot prove.
Published methodology
A reviewable audit method, not a black-box score
See what OpenForBots requests from a public website, how it interprets the response, why a finding appears, and which questions still require human review or evidence from an external provider.
How to read an audit
Evidence, interpretation, action, and verification stay separate
The report shows the public source behind a finding, explains what that evidence supports, recommends a practical next step, and states what a later check should confirm.
- Website scope
- Public hostname and selected standard public resources
- Evidence labels
- Observed, documented, simulated, heuristic, inferred, or manual review
- Report contract
- ofb-audit-r2.1
- Evidence engine
- ofb-evidence-r1-2026-08-02
- Roadmap rules
- ofb-roadmap-r2-2026-08-03
- Report storage
- Anonymous audit snapshots retained for up to 30 days
- Latest content update
- Information checked
Method documents
Inspect the reasoning behind the report
Each document explains its sources, review date, practical limits, and the part of the audit it helps you interpret.