24h AI reads
55
14 agent-readable route reads.
Experiment journal · daily run
Agent-readability experiment log
This page keeps crawler telemetry and answerability benchmark signals separate. Bot reads prove access. The benchmark checks whether profiles contain enough facts for agent-style tasks.
7d AI reads
83
25 agent-readable route reads.
Benchmark avg
85.8/100
250 profiles scored by deterministic completeness checks.
Top AI bots · 7d
- ClaudeBot: 41
- OAI-SearchBot: 23
- GPTBot: 12
- ChatGPT-User: 7
Top agent-readable paths · 7d
/sitemap-agents.xml: 16/docs/agent-readability-log/: 2/.well-known/agent-card.json: 1/.well-known/agent-skills/index.json: 1/.well-known/agenthub.json: 1/.well-known/api-catalog: 1/articles/product-layer-for-ai-agents-readable-offers.html: 1/company-profile.json: 1/llms.txt: 1
Status breakdown
- 200: 24
- 404: 1
The goal is not more bot traffic at any cost. The first goal is stable 2xx access to agent-readable routes.
Last experiment reads
- 2026-04-27T21:51:50.000Z · GPTBot · 200 ·
/.well-known/agent-skills/index.json - 2026-04-27T23:14:02.000Z · ClaudeBot · 200 ·
/sitemap-agents.xml - 2026-04-28T02:26:44.000Z · ClaudeBot · 200 ·
/sitemap-agents.xml - 2026-04-28T05:06:00.000Z · ClaudeBot · 200 ·
/sitemap-agents.xml - 2026-04-28T08:33:37.000Z · ClaudeBot · 200 ·
/sitemap-agents.xml - 2026-04-28T11:26:17.000Z · ClaudeBot · 200 ·
/sitemap-agents.xml - 2026-04-28T14:04:07.000Z · ClaudeBot · 200 ·
/sitemap-agents.xml - 2026-04-28T15:37:33.000Z · GPTBot · 200 ·
/sitemap-agents.xml - 2026-04-28T16:15:10.000Z · GPTBot · 404 ·
/docs/agent-readability-log/ - 2026-04-28T16:19:28.000Z · GPTBot · 200 ·
/docs/agent-readability-log/
Deterministic answerability benchmark
First 250 local public hub profiles, scored by deterministic completeness checks. This is not proof of ranking uplift or AI recommendation inclusion. It measures whether profiles contain enough structured facts for agent-style tasks.
| Cohort | Profiles | Average score |
|---|---|---|
| D agent guidance | 178 | 80/100 |
| E super profile | 72 | 100/100 |
What this proves
- AI bots can reach the public agent-readable layer.
- Status codes show whether the layer is operational.
- Profile completeness can be scored consistently over time.
What this does not prove
- No ranking promise.
- No guarantee of AI recommendations.
- No proof of buyer conversion without funnel data.