ai-readiness-check
Skill0 starsChanged 29 days agoMITView on GitHub
Use for AI search readiness and AI crawler access, for example "can ChatGPT read my site", "is my site blocked for AI", "check GPTBot, ClaudeBot, PerplexityBot or Google-Extended", "AI visibility check", "is my robots.txt blocking AI", "do I need llms.txt", or "GEO readiness". Checks each AI search and user-fetch bot against the live site, runs the readiness checklist, and reports every bot as ok or blocked with the reason.
In this repository10 other files it gives its agents.
- analytics-ga4Skill
- approve-and-measure-changeSkill
- connect-and-importSkill
- core-web-vitalsSkill
- leadsSkill
- pages-to-winSkill
- search-consoleSkill
- site-auditSkill
- start-hereSkill
- tag-manager-trackingSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 3 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 16
Every bot in the result is reported as ok or blocked, each with its reason (the robots.txt rule, the HTTP status, or the firewall response).
- Distinctive: line 17
Keep the two bot kinds apart: search bots (which index pages for AI answers) and user-fetch bots (which fetch a page when a person asks …
- Distinctive: line 18
Never state a rank or position inside any AI assistant, and never claim the site "appears in ChatGPT".
- Distinctive: line 19
Never scrape or query consumer AI apps to test visibility.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
AI readiness check
Why: an assistant cannot cite a page its crawler is not allowed to fetch. This is the cheapest check with the biggest effect, so it runs before any content work.
Rules that come first
- This skill only reads. It never edits robots.txt, llms.txt or any page. A fix goes through approve-and-measure-change.
- Every bot in the result is reported as
okorblocked, each with its reason (the robots.txt rule, the HTTP status, or the firewall response). No bot is left out and none is reported as "probably fine". - Keep the two bot kinds apart: search bots (which index pages for AI answers) and user-fetch bots (which fetch a page when a person asks about it). Blocking one kind is a different decision from blocking the other.
- Never state a rank or position inside any AI assistant, and never claim the site "appears in ChatGPT". This check measures access, not visibility.
- Never scrape or query consumer AI apps to test visibility.
- Any rate you report comes with its n plus an interval, or the words "insufficient data".
Steps
- Get the site and its domain from start-here (
gravity_overview, thengravity_list_sitesif needed). gravity_geo_crawler_accessfor the domain: the per-bot result for AI search and user-fetch bots.gravity_seo_checklist_runfor the site: the readiness checklist. Read the item list withgravity_seo_checklist_itemswhen you need the wording of an item.- Optional reads when the user asks for more depth:
gravity_geo_llmstxt_validateto check an existing llms.txt.gravity_geo_fetch_pageto see a page the way a bot fetches it.gravity_geo_citabilityto see whether a page is easy to quote.
- Report:
- A table with one row per bot: bot, kind (search or user-fetch),
okorblocked, reason. - The checklist items that failed, each with what to change and where.
- One next step. If a bot the user cares about is blocked, offer the fix through approve-and-measure-change.
- A table with one row per bot: bot, kind (search or user-fetch),
Check
The lead's eval proves this skill when the reply calls gravity_geo_crawler_access and gravity_seo_checklist_run, contains a row for every bot the tool returned with the literal ok or blocked and a reason, and contains no AI rank position and no claim of being shown in an assistant's answers.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

