R1 High confidence Proposed

Open approved AI crawler access

One receipt from evidence through decision, application, verification, and outcome.

Priority 34.20 impact 38.0 × confidence 90% ÷ effort 1

Decision receipt

Why this fires

Sample evidence

The attached observations crossed this rule's declared threshold. Fixing it changes the named target; measurement starts only after verification.

Page readiness audit 2026-09-13

Target page: https://larkspur-roasters.example/
Target query: —
Impact estimate: 38.0
Measurement window: 14 days after verification

1Page readiness audit {"check": "robots", "page": "/"}2026-09-13 22:10

Options considered

Choose the road, not just the recommendation.

Implementation handoff

Artifact

robots diff Preview · not saved

For the approach: Allow selected search and answer-engine crawlers while keeping training opt-outs explicit.

--- robots.txt
+++ robots.txt
@@ AI crawler access @@
+User-agent: ChatGPT-User
+Allow: /
+User-agent: OAI-SearchBot
+Allow: /
+User-agent: PerplexityBot
+Allow: /
+Content-Signal: search=yes, ai-input=yes, ai-train=no
+# CiteSonar-Verify: gs-robots-diff-d57d9a01f2e95952849725bf3689ff45

The classic User-agent rules allow search and user-agent retrieval. The
Content-Signal line expresses the same search/AI-input policy; ai-train keeps
the licensing stance already visible in the audited robots policy. Review that
training choice before publishing.
The 3 operators
above are exactly the
search-and-agent crawlers your latest
readiness audit found blocked. Operators you block deliberately for training-only
access are left untouched.

Verification marker# CiteSonar-Verify: gs-robots-diff-d57d9a01f2e95952849725bf3689ff45

Rendered by the same deterministic generator a workspace uses. There, accepting issues this file, and marking it applied starts a crawl that looks for the marker before anything is measured. Create a workspace for your site

Before and after

Outcome

An outcome appears after the verified measurement window closes.