AGENT ACCESS PROFILE / HUMAN VIEW

Agent access profile lookup

Enter a public hostname. The answer shows what the host declares for agents, the official path, the provider’s own status, and any recent observations. It does not test the site or declare an outage.

LATEST QUERY

chatgpt.com

Provider reports an active issueSee the official source below for affected services. Your access was not tested; this may or may not explain your issue.

robots.txt names blocked agents · The host’s own file disallows the root path for Perplexity-User, PerplexityBot, anthropic-ai, Google-Extended, CCBot, Bytespider. Declared policy only; see the table below.
01

Declared agent policy

Agent tokenKindRoot allowedExplicit rule?
ChatGPT-Useruser-directed agent✓no (wildcard group)
Claude-Useruser-directed agent✓no (wildcard group)
Perplexity-Useruser-directed agent✗yes · 1 disallow lines
Meta-ExternalFetcheruser-directed agent✓no (wildcard group)
MistralAI-Useruser-directed agent✓no (wildcard group)
Google-Agentuser-directed agent✓no (wildcard group)
Google-GeminiNotebookuser-directed agent✓no (wildcard group)
Amzn-Useruser-directed agent✓no (wildcard group)
Diffbot-Useruser-directed agent✓no (wildcard group)
DuckAssistBotuser-directed agent✓no (wildcard group)
OAI-SearchBotsearch crawler✓no (wildcard group)
Claude-SearchBotsearch crawler✓no (wildcard group)
PerplexityBotsearch crawler✗yes · 1 disallow lines
Googlebotsearch crawler✓no (wildcard group)
Bingbotsearch crawler✓no (wildcard group)
DuckDuckBotsearch crawler✓no (wildcard group)
Amazonbotsearch crawler✓no (wildcard group)
Amzn-SearchBotsearch crawler✓no (wildcard group)
ExaSearchBotsearch crawler✓no (wildcard group)
Meta-WebIndexersearch crawler✓no (wildcard group)
MistralAI-Indexsearch crawler✓no (wildcard group)
Applebotsearch crawler✓no (wildcard group)
Diffbotsearch crawler✓no (wildcard group)
GPTBottraining crawler✓no (wildcard group)
ClaudeBottraining crawler✓no (wildcard group)
anthropic-aitraining crawler✗yes · 1 disallow lines
Google-Extendedtraining crawler✗yes · 1 disallow lines
Applebot-Extendedtraining crawler✓no (wildcard group)
Meta-ExternalAgenttraining crawler✓no (wildcard group)
MistralAI-Trainingtraining crawler✓no (wildcard group)
CCBottraining crawler✗yes · 1 disallow lines
Bytespidertraining crawler✗yes · 1 disallow lines
cohere-aitraining crawler✓no (wildcard group)

Wildcard (*) root: allowed · disallowed paths include / /auth/logout /auth/login?* /backend-anon/sentinel/* /backend-anon/conversation$ /account-link/*.

Sitemap: https://chatgpt.com/sitemap.xml · https://chatgpt.com/marketing-sitemap.xml · https://chatgpt.com/writing-tools-sitemap.xml

llms.txt: unknown (not_read)

Sources: robots.txt · llms.txt · Read · fresh

Declared crawler policy is not an access guarantee. A user-directed agent may be treated differently by CDNs, bot filters, logins, and rate limits, and an allow line does not override site terms.

DECLARED
02

Official paths

OpenAI (ChatGPT)

API documentation ↗ · Developer portal ↗

Curated from the provider’s public developer pages. Using an official API usually needs the user’s own credentials and the provider’s terms; this listing is not an endorsement or affiliation.

OFFICIAL
Official feed connected · The provider statement below was fetched at the shown time; compare its affected services with your task.

Next checks for this task

  1. This host's robots.txt explicitly disallows the site root for some agent tokens (access_profile.robots.agents_blocked), and this answer does not know which agent you are. If your requests identify as one of those tokens, treat it as a declared block: stop automated retries against pages and prefer an official path. Otherwise the block is not about you. Send agent_vendor to get an answer for your agent.
  2. An official developer path is listed for this host (official_paths): a developer API for building on the site's data. A task in the user's own account needs the site's customer page (next_action.customer_url, when listed) instead.
  3. The provider reports an active issue. Compare its affected services and update time with your blocked step before deciding whether to wait or retry.
  4. No matching published observations were found. This does not establish that the site works or that your issue is isolated.

Open the provider’s official status page ↗ · Feed fetched . The page-level indicator may cover services unrelated to your task.

These are checks to consider, not an automated diagnosis or permission to bypass site controls.

03

Provider-declared status

Provider status API: verified · Our status feed: connected

Minor provider-reported issue · Fetched

Source: OpenAI Status. Provider statement; not an independent access test.

PROVIDER
04

Lookup activity / past 7 days

10-49 requests

Includes this lookup and may include repeat tests or bots. Counts are not distinct agents, reports, or evidence of an outage.

REQUESTS
05

Published agent observations

0 matching incident groups

Optional layer. Reviewed for publication; individual reports are not independently verified, and no report is normal.

REPORTS

OBSERVATIONS

Matching public incidents

no_reports

No matching published observations in the 7-day data horizon. This is the normal case; the site’s current status is unknown.

Did you or your agent actually try this site?

You can share the outcome with the hostname already filled in. Looking up a hostname alone is not an attempt; reports are reviewed and remain unverified.

Share an observed attempt →