CrawlLedger
Evidence brief · generated from public metadata

Evidence briefs / cloudflare.com

cloudflare.com policy evidence brief.

A bounded report of what CrawlLedger observed, when it observed it, how the versioned parser classified it, and which public hashes identify the retained artifacts.

As of Sep 23, 2026

Generated on request from append-only metadata. Refreshing this page does not crawl the domain.

Executive evidence summary

These counts describe the bounded CrawlLedger record for this domain. They do not establish ownership, intent, crawler identity, compliance, contract, or legal permission.

86observations
2signal types
0parsed changes
Aug 4, 2026first observed

Current observation register

The artifact SHA-256 identifies retained bytes without exposing a private storage location or redistributing the source body.

SignalObservedHTTPStateArtifact SHA-256Parser
llms.txt301Unverified after HTTP responseNo retained artifactNot applied
robots.txt200Observed and parsedc6e360cc6e931ea35c33ae917ff9e2194eb55f79f68e949ec54bcd4533324d98robots-rfc9309-rsl-content-signals-0.2.0

Current robots.txt interpretation

Each state uses the registry and parser version attached to the observation. “Allowed” is a parser result, not a grant or verified crawler behavior.

Tracked crawler observations

These are parser results for the published robots rules, not authentication or legal conclusions.

Crawler tokenObserved stateMatched group
GPTBot
OpenAI · training
allowedGPTBot
OAI-SearchBot
OpenAI · search
allowed*
OAI-AdsBot
OpenAI · ads_validation
unverifiedNo matching group observed
ChatGPT-User
OpenAI · user_action
allowedChatGPT-User
ClaudeBot
Anthropic · training
allowed*
Claude-SearchBot
Anthropic · search
allowed*
Claude-User
Anthropic · user_action
allowed*
Google-Extended
Google · mixed
allowedGoogle-Extended
Applebot-Extended
Apple · training
allowed*
PerplexityBot
Perplexity · search
allowedPerplexityBot
Perplexity-User
Perplexity · user_action
allowed*
CCBot
Common Crawl · archive
allowedCCBot

Observed Content-Signal declarations

These emerging robots.txt extensions state publisher preferences about downstream use. CrawlLedger records the declarations without treating them as enforcement or legal permission.

User-agent groupPathSearchAI inputAI trainingReuse
Cohere-aiall pathsyesyesyesnot stated

Learn how Content Signals differ from crawl rules →

Recent append-only timeline

Chain identifiers make replacement or branching detectable. A conditional 304 can extend the chain while reusing the prior artifact hash.

ObservedSignalStateChangedChain hashPrevious chain
llms.txtUnverified after HTTP responseNoe69ebb8450a1ce2cd38d03f9ea1985aa9135e03de58f5ff178f9389374503882c1a681221de81417c76c14cee8e683bc4f318c6de4cc73167ac1acc1120eed87
robots.txtObserved and parsedNo35571f5eb6a9e8ce51b8c808512d3b8dff8e8b58ac36d9bf361018386a2ee1f0adbeda01f79393c4e79a17aa7983956e74fa7e8df42bf89b9a3fdfeb0a38b563
llms.txtUnverified after HTTP responseNoc1a681221de81417c76c14cee8e683bc4f318c6de4cc73167ac1acc1120eed8717274721d78dc035ce1eb3bf6cd7061bd47c447326c5ee4ac1378144d264f71f
robots.txtObserved and parsedNoadbeda01f79393c4e79a17aa7983956e74fa7e8df42bf89b9a3fdfeb0a38b56341064c163dbe03e0a1b1de99e11604109f58eb2127756e42d17aafb6b04d475a
llms.txtUnverified after HTTP responseNo17274721d78dc035ce1eb3bf6cd7061bd47c447326c5ee4ac1378144d264f71f698ad07bd78546cb987e9b5dd856b4a260f7a860d616b4a1257506b732e611c2
robots.txtObserved and parsedNo41064c163dbe03e0a1b1de99e11604109f58eb2127756e42d17aafb6b04d475a60aaa705c7eedd89a5f512d84bc72ff00e0bd77233df417e184e50dcfcf973fe
llms.txtUnverified after HTTP responseNo698ad07bd78546cb987e9b5dd856b4a260f7a860d616b4a1257506b732e611c2157fe6bf5575bc548939d5032e68322851b9772cd2d5eca67c2a8db9d20b63d2
robots.txtObserved and parsedNo60aaa705c7eedd89a5f512d84bc72ff00e0bd77233df417e184e50dcfcf973fe0a77a65e1e9c7be8f575ddd61497d491d35b72ab6d32495851a63205b5a09dea
llms.txtUnverified after HTTP responseNo157fe6bf5575bc548939d5032e68322851b9772cd2d5eca67c2a8db9d20b63d2cdabff998cda72ff74020dbec42263f2b4bbefac478df9048d6be4851060f1b2
robots.txtObserved and parsedNo0a77a65e1e9c7be8f575ddd61497d491d35b72ab6d32495851a63205b5a09deac3bb9f775ab053f22197666419a8e537208c60748a155a8682f0bfd102fa719c
llms.txtUnverified after HTTP responseNocdabff998cda72ff74020dbec42263f2b4bbefac478df9048d6be4851060f1b2b0d9f15f33230ad90504b5f60ac8a6725470bc99153652ed00f2296ad2828998
robots.txtObserved and parsedNoc3bb9f775ab053f22197666419a8e537208c60748a155a8682f0bfd102fa719c7d4712173be63cdbe7a308caca159bc654ee349fb7b57b99831bb174731a2fda
llms.txtUnverified after HTTP responseNob0d9f15f33230ad90504b5f60ac8a6725470bc99153652ed00f2296ad28289986918c0a57be755996feced295e8aed70b6b6b615bf19ec2f108b70137e227b73
robots.txtObserved and parsedNo7d4712173be63cdbe7a308caca159bc654ee349fb7b57b99831bb174731a2fda3ad1a4c88fb4711ae62f7548d491171ee1cac6d610c0d9e4a4006e595321005b
llms.txtUnverified after HTTP responseNo6918c0a57be755996feced295e8aed70b6b6b615bf19ec2f108b70137e227b73e63b193f8b29cb86f4f8896c9723a6806b0bc5b7f96259650837eb13ef79df9a
robots.txtObserved and parsedNo3ad1a4c88fb4711ae62f7548d491171ee1cac6d610c0d9e4a4006e595321005b73cbda53ac124d46bf38dc63e5384f7078528055f614f61f1fecea2bb7205ec0
llms.txtUnverified after HTTP responseNoe63b193f8b29cb86f4f8896c9723a6806b0bc5b7f96259650837eb13ef79df9a218657bc50eb35f490516d7ad477c9ed8d6d1f6928a6844fd49ac33a37ab5a1a
robots.txtObserved and parsedNo73cbda53ac124d46bf38dc63e5384f7078528055f614f61f1fecea2bb7205ec030b8eadda1eca58b8b5f38422448301ccaf4d6c557a05cbf70b14dd3d76e62e7
llms.txtUnverified after HTTP responseNo218657bc50eb35f490516d7ad477c9ed8d6d1f6928a6844fd49ac33a37ab5a1a168aa7391ec4bf30ca8f395dcb8e1bf68b7ab0419acb2de591c139bb8a2230bf
robots.txtObserved and parsedNo30b8eadda1eca58b8b5f38422448301ccaf4d6c557a05cbf70b14dd3d76e62e73305713d984510eae54a65c0f2c91f15713922851f5bada0be27cba877c70a45

Method and limitations

CrawlLedger observes narrow public policy endpoints, records bounded response metadata, retains source bytes privately by content hash, and appends parser output. It does not authenticate requesting bots, infer unstated intent, or decide legal permission.

Included

Public metadata, exact hashes, parser versions, signal states, and recent chain links from the fixed cohort.

Not included

Private artifact keys, raw source redistribution, ordinary page content, legal conclusions, or customer-specific monitoring.

How this report was produced: CrawlLedger assembled it automatically from append-only observation metadata and a versioned parser registry. Documentation and product boundaries were source-reviewed on Sep 20, 2026. No new network fetch, customer data, or legal interpretation was added for this export.