Agent Lens: AI Readiness & Markdown

Agent Lens helps web developers inspect how a website presents itself to AI agents and read the text/markdown representation served by that website. Open the compact side panel on any HTTP or HTTPS page to run an on-demand, local agent-readiness check. Agent Lens reports a score, maturity level, coverage, evidence, and practical implementation guidance. Public signals checked include: - robots.txt, XML sitemaps, HTTP Link headers, and DNS-AID - Accept: text/markdown content negotiation - AI crawler rules, Content Signals, and Web Bot Auth - API Catalog, MCP Server Cards, Agent Skills, and WebMCP - OAuth and OIDC discovery, Protected Resource Metadata, and auth.md - Optional x402, MPP, UCP, and ACP commerce signals Markdown View requests the current URL with Accept: text/markdown. Read the response as full-width raw text or sanitized, rendered GitHub Flavored Markdown. Follow web links while staying in the viewer, use the table of contents, inspect the final source URL, copy or download content, switch themes, control remote images, view a character count and local token estimate, and see leading front matter unchanged in a text information card. Privacy by design: - Analysis, scoring, and reports stay in your browser - No account, analytics, telemetry, advertising, or publisher-operated scanning server - Scan and document requests omit website cookies and Authorization headers - Website access is requested for the selected origin only after a user action - Discovered WebMCP tools are inspected but never executed - No remotely hosted executable code A full scan uses Cloudflare's DNS-over-HTTPS service to check DNS-AID, which sends the checked domain name to that resolver. If image loading is enabled, images are requested directly from the hosts referenced by the document. Preferences and compact reports are stored locally and can be deleted with the extension. Agent Lens is built for web developers, technical SEO specialists, API platform teams, and AI infrastructure teams that need to validate agent-ready website signals or debug content negotiation without sending reports to a third-party scanning service. Results are client-side observations from your Chrome session. CDN edges, WAF rules, network location, and experimental browser APIs can produce different results. Finding discovery metadata does not by itself prove complete protocol conformance.