Changes that affect what you can inspect or trust

Bot-range verification, fuller response comparison and privacy-safe measurement

Added the missing bridge between a claimed crawler name in a log and evidence published by the operator.

• The new bilingual Bot Request Verifier checks one local log line against its exact product token and dated Google, OpenAI, Perplexity or Anthropic IP ranges without uploading the line or address.
• Crawler response comparisons now include a bounded response-header allowlist alongside status, canonical, noindex, visible text and body hash.
• Cookie-free Umami measurement records only allowlisted navigation, tool-state and report-action categories; URLs, hosts, IPs, logs and pasted evidence are excluded.

Visible crawler policy, portable reports and a controlled second view

Moved the crawler-purpose matrix into the main result, added local evidence export and introduced a bounded comparison request for one selected crawler product token.

• Search, training and user-triggered fetch roles remain separate and link to current operator sources.
• Up to five reports can remain in this browser for 24 hours for a before/after comparison; Markdown copy and JSON export stay local.
• The second view is labelled as a simulated header, not an authenticated Google, OpenAI, Anthropic or Perplexity crawler.
• The crawler lab now publishes the first controlled response baseline and its raw JSON evidence.
• Two bilingual, primary-source articles now separate robots policy from CDN enforcement and initial HTML from the JavaScript-rendered DOM.
• The German homepage routes Redirect, Robots and Server Log actions to their maintained German tools.
• Privacy and methodology now distinguish disabled client-side audience analytics from limited gateway operations metrics and warn against signed or secret-bearing URLs.

Shorter tool choices and clearer result handoffs

Made the specialist tools easier to scan and move between. Results now open in view instead of appearing unnoticed below the form, and long guides use a tighter reading rhythm.

• The tool index uses five compact cards instead of tall presentation blocks.
• Every specialist tool links directly to the other focused checks.
• Successful live checks may be reused by the gateway for up to one hour; failures for five minutes. A later release added the separately documented 24-hour browser-local comparison history.

The live crawler check is now the main product

Rebuilt the homepage around the fast one-URL check, practical follow-up questions, a role-based crawler study and in-depth learning paths. Pale topic bands and the footer now extend to the viewport edge while their content remains readable.

• Expanded the crawler comparison to eight distinct search, training, user-fetch and control roles.
• Published a bilingual primary-source guide to technical eligibility for AI search.
• Expanded the three former short lab notes into full technical SEO articles and removed the internal label from the visible interface.
• Raised the protected allowance to six checks per ten minutes and 30 per 24 hours without increasing concurrency or worker CPU.

One protected crawler check with four plain answers

Added a bilingual live URL check for fetch access, search-crawler access, indexability, and server-visible main content. The result exposes technical evidence without turning it into a synthetic visibility score.

• The operation uses one concurrent slot, 30 runs per 24 hours per pseudonymous client, and no paid SEO or LLM provider.
• The public-tools container is capped at 0.5 CPU and 256 MB, separate from the Crawl Foundry Site Audit worker.
• Privacy, methodology, sitemap, CSP, internal links, and common-owner disclosures now match the live data path.

Simpler design, bilingual entry pages and crawler comparison

Reorganised the site around three clear jobs: use a tool, understand a crawling question or compare crawler systems. Replaced the heavy display style with a readable system sans-serif and a lighter visual hierarchy.

• Added German entry pages for the home, tools, crawler and knowledge areas.
• Added a primary-source comparison of search and AI crawler roles without naming an unsupported product winner.
• Related Fan-out links appear only in relevant sections with common ownership visible; no Crawl Foundry product link was added.

Redirect chain evidence and crawler verification comparison

Replaced the interim redirect-chain alias with a dedicated browser-local chain builder, executable parsing fixtures, and structured evidence output. Added a primary-source comparison of crawler verification methods.

• Relative Location fields, capture gaps, loops, and incomplete final states stay visible.
• No tool fetches a URL or verifies a crawler identity on the visitor's behalf.
• The log inspector can now export aggregate evidence without the raw log lines.

Robots rule testing and executed fixtures

Added a browser-local RFC 9309 rule tester, a reproducible testing guide, and seven versioned fixtures whose expected decisions run during build verification.

• The tool performs no network request and does not claim crawler-vendor parity.
• The benchmark page now publishes actual local fixture outcomes instead of a protocol-only placeholder.
• The repository records one primary job for every indexable canonical URL.

Canonical pages released for indexing

Removed the coordinated metadata, response-header, and robots blocking signals after the owner-approved minimum viable launch passed legal, legacy URL, sitemap, responsive, and production checks.

• The sitemap is generated from the canonical route and content registries and contains only indexable 200 pages.
• The `www` host permanently preserves paths while redirecting to the canonical apex.
• Tool, guide, reference, and article schema now describes the visible page role.

Operator and privacy boundary published

Added the verified legal operator details and a technology-specific privacy notice covering Vercel request delivery, browser-local tool processing, email contact, external links, and data-subject rights.

• No client-side audience analytics, advertising tags, accounts, or contact form are currently active; limited gateway operations and abuse-prevention metrics are documented separately.
• Tool input and results stay in the visitor's browser; hosting request metadata remains a separate infrastructure boundary.
• The acquired-domain ownership and rights separation remains explicit.

Diagnostic library and responsive type hardening

Added dedicated Tools, Guides, Lab Notes, Reference, and audience hubs. Published four diagnostic guides, three primary-source reference pages, three dated lab notes, About, and this changelog.

• Large headings wrap inside their grid track instead of being clipped by fixed sidebar widths or no-wrap rules.
• Navigation, content rows, tables, source links, and technical strings have explicit narrow-screen behavior.
• This release remained noindex until the later public-launch gate passed.

Three browser-local inspectors

Introduced a bounded access-log parser, a paste-only HTTP response and HTML inspector, and a local IP syntax and reserved-range check.

• No current tool uploads submitted data.
• No current tool fetches an arbitrary URL or scans a target.
• Results distinguish observations, inferences, and unknowns.

New ownership and legacy URL decisions

Published the new-ownership boundary and URL-level decisions for the strongest historical targets. No former software, content, customers, or operator identity is presented as transferred.