afdocs 0.1.3 → 0.2.0

package/dist/helpers/get-page-urls.js

@@ -0,0 +1,230 @@
+import { extractMarkdownLinks } from '../checks/llms-txt/llms-txt-valid.js';
+import { MAX_SITEMAP_URLS } from '../constants.js';
+/**
+ * Extract `<loc>` URLs from a sitemap XML string.
+ * Also returns any `<sitemap><loc>` entries from sitemap index files.
+ */
+export function parseSitemapUrls(xml) {
+    const urls = [];
+    const sitemapIndexUrls = [];
+    // Detect sitemap index: contains <sitemap> elements
+    const isSitemapIndex = /<sitemap[\s>]/i.test(xml);
+    if (isSitemapIndex) {
+        // Extract <loc> inside <sitemap> blocks
+        const sitemapBlockRegex = /<sitemap[\s>][\s\S]*?<\/sitemap>/gi;
+        let block;
+        while ((block = sitemapBlockRegex.exec(xml)) !== null) {
+            const locMatch = /<loc>\s*(.*?)\s*<\/loc>/i.exec(block[0]);
+            if (locMatch) {
+                sitemapIndexUrls.push(locMatch[1]);
+            }
+        }
+    }
+    else {
+        // Regular sitemap: extract all <loc> entries
+        const locRegex = /<loc>\s*(.*?)\s*<\/loc>/gi;
+        let match;
+        while ((match = locRegex.exec(xml)) !== null) {
+            urls.push(match[1]);
+        }
+    }
+    return { urls, sitemapIndexUrls };
+}
+function getUrlsFromCachedLlmsTxt(ctx) {
+    const existsResult = ctx.previousResults.get('llms-txt-exists');
+    const discovered = (existsResult?.details?.discoveredFiles ?? []);
+    return extractLinksFromLlmsTxtFiles(discovered);
+}
+function extractLinksFromLlmsTxtFiles(files) {
+    const urls = new Set();
+    for (const file of files) {
+        const links = extractMarkdownLinks(file.content);
+        for (const link of links) {
+            if (link.url.startsWith('http://') || link.url.startsWith('https://')) {
+                urls.add(link.url);
+            }
+        }
+    }
+    return Array.from(urls);
+}
+/**
+ * Directly fetch llms.txt candidate URLs and extract links.
+ * Used when `llms-txt-exists` hasn't run (e.g. standalone check mode).
+ */
+async function fetchLlmsTxtUrls(ctx) {
+    const candidates = new Set();
+    candidates.add(`${ctx.baseUrl}/llms.txt`);
+    candidates.add(`${ctx.origin}/llms.txt`);
+    candidates.add(`${ctx.origin}/docs/llms.txt`);
+    const discovered = [];
+    for (const url of candidates) {
+        try {
+            const response = await ctx.http.fetch(url);
+            if (!response.ok)
+                continue;
+            const contentType = response.headers.get('content-type') ?? '';
+            if (!contentType.includes('text/'))
+                continue;
+            const content = await response.text();
+            const trimmed = content.trimStart().toLowerCase();
+            if (trimmed.startsWith('<!') || trimmed.startsWith('<html'))
+                continue;
+            if (content.trim().length === 0)
+                continue;
+            discovered.push({
+                url,
+                content,
+                status: response.status,
+                redirected: response.redirected,
+            });
+        }
+        catch {
+            // Skip failed fetches
+        }
+    }
+    return extractLinksFromLlmsTxtFiles(discovered);
+}
+/**
+ * Parse `Sitemap:` directives from a robots.txt body.
+ * Returns an array of sitemap URLs found.
+ */
+export function parseSitemapDirectives(robotsTxt) {
+    const urls = [];
+    for (const line of robotsTxt.split('\n')) {
+        const match = /^\s*Sitemap:\s*(\S+)/i.exec(line);
+        if (match) {
+            urls.push(match[1]);
+        }
+    }
+    return urls;
+}
+/**
+ * Discover sitemap URLs by checking robots.txt first, then falling back to /sitemap.xml.
+ */
+async function discoverSitemapUrls(ctx) {
+    // Try robots.txt for Sitemap directives
+    try {
+        const robotsResponse = await ctx.http.fetch(`${ctx.origin}/robots.txt`);
+        if (robotsResponse.ok) {
+            const body = await robotsResponse.text();
+            const directives = parseSitemapDirectives(body);
+            if (directives.length > 0)
+                return directives;
+        }
+    }
+    catch {
+        // robots.txt fetch failed; fall through
+    }
+    // Default to /sitemap.xml
+    return [`${ctx.origin}/sitemap.xml`];
+}
+function isGzipped(url) {
+    return /\.gz($|\?)/i.test(url);
+}
+async function fetchSitemap(ctx, sitemapUrl, warnings) {
+    if (isGzipped(sitemapUrl)) {
+        warnings.push(`Skipped gzipped sitemap (not supported): ${sitemapUrl}`);
+        return { urls: [], sitemapIndexUrls: [] };
+    }
+    try {
+        const response = await ctx.http.fetch(sitemapUrl);
+        if (!response.ok)
+            return { urls: [], sitemapIndexUrls: [] };
+        const xml = await response.text();
+        return parseSitemapUrls(xml);
+    }
+    catch {
+        return { urls: [], sitemapIndexUrls: [] };
+    }
+}
+async function getUrlsFromSitemap(ctx, warnings) {
+    const sitemapUrls = await discoverSitemapUrls(ctx);
+    const urls = [];
+    for (const sitemapUrl of sitemapUrls) {
+        if (urls.length >= MAX_SITEMAP_URLS)
+            break;
+        const parsed = await fetchSitemap(ctx, sitemapUrl, warnings);
+        // Add direct URLs (filtered to same origin)
+        for (const url of parsed.urls) {
+            if (urls.length >= MAX_SITEMAP_URLS)
+                break;
+            try {
+                const u = new URL(url);
+                if (u.origin === ctx.origin) {
+                    urls.push(url);
+                }
+            }
+            catch {
+                // Skip malformed URLs
+            }
+        }
+        // Follow one level of sitemap index
+        if (parsed.sitemapIndexUrls.length > 0 && urls.length < MAX_SITEMAP_URLS) {
+            for (const subSitemapUrl of parsed.sitemapIndexUrls) {
+                if (urls.length >= MAX_SITEMAP_URLS)
+                    break;
+                const subParsed = await fetchSitemap(ctx, subSitemapUrl, warnings);
+                for (const url of subParsed.urls) {
+                    if (urls.length >= MAX_SITEMAP_URLS)
+                        break;
+                    try {
+                        const u = new URL(url);
+                        if (u.origin === ctx.origin) {
+                            urls.push(url);
+                        }
+                    }
+                    catch {
+                        // Skip malformed URLs
+                    }
+                }
+            }
+        }
+    }
+    return urls;
+}
+/**
+ * Discover page URLs from llms.txt links, sitemap, or fall back to baseUrl.
+ *
+ * Priority:
+ * 1. llms.txt links (from previous check results)
+ * 2. Sitemap URLs (robots.txt Sitemap directives, then /sitemap.xml fallback)
+ * 3. baseUrl fallback
+ */
+export async function getPageUrls(ctx) {
+    const warnings = [];
+    // 1. Try llms.txt links from cached results (if llms-txt-exists ran)
+    const cachedUrls = getUrlsFromCachedLlmsTxt(ctx);
+    if (cachedUrls.length > 0)
+        return { urls: cachedUrls, warnings };
+    // 2. Try fetching llms.txt directly (standalone mode, llms-txt-exists didn't run)
+    if (!ctx.previousResults.has('llms-txt-exists')) {
+        const fetchedUrls = await fetchLlmsTxtUrls(ctx);
+        if (fetchedUrls.length > 0)
+            return { urls: fetchedUrls, warnings };
+    }
+    // 3. Try sitemap
+    const sitemapUrls = await getUrlsFromSitemap(ctx, warnings);
+    if (sitemapUrls.length > 0)
+        return { urls: sitemapUrls, warnings };
+    // 4. Fallback
+    return { urls: [ctx.baseUrl], warnings };
+}
+/**
+ * Discover page URLs and sample down to maxLinksToTest if needed.
+ * Consolidates the discover+shuffle+sample pattern used by multiple checks.
+ */
+export async function discoverAndSamplePages(ctx) {
+    const discovery = await getPageUrls(ctx);
+    let urls = discovery.urls;
+    const totalPages = urls.length;
+    const sampled = totalPages > ctx.options.maxLinksToTest;
+    if (sampled) {
+        for (let i = urls.length - 1; i > 0; i--) {
+            const j = Math.floor(Math.random() * (i + 1));
+            [urls[i], urls[j]] = [urls[j], urls[i]];
+        }
+        urls = urls.slice(0, ctx.options.maxLinksToTest);
+    }
+    return { urls, totalPages, sampled, warnings: discovery.warnings };
+}
+//# sourceMappingURL=get-page-urls.js.map

package/dist/helpers/get-page-urls.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"get-page-urls.js","sourceRoot":"","sources":["../../src/helpers/get-page-urls.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,oBAAoB,EAAE,MAAM,sCAAsC,CAAC;AAC5E,OAAO,EAAE,gBAAgB,EAAE,MAAM,iBAAiB,CAAC;AAGnD;;;GAGG;AACH,MAAM,UAAU,gBAAgB,CAAC,GAAW;IAC1C,MAAM,IAAI,GAAa,EAAE,CAAC;IAC1B,MAAM,gBAAgB,GAAa,EAAE,CAAC;IAEtC,oDAAoD;IACpD,MAAM,cAAc,GAAG,gBAAgB,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IAElD,IAAI,cAAc,EAAE,CAAC;QACnB,wCAAwC;QACxC,MAAM,iBAAiB,GAAG,oCAAoC,CAAC;QAC/D,IAAI,KAAK,CAAC;QACV,OAAO,CAAC,KAAK,GAAG,iBAAiB,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;YACtD,MAAM,QAAQ,GAAG,0BAA0B,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC;YAC3D,IAAI,QAAQ,EAAE,CAAC;gBACb,gBAAgB,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC;YACrC,CAAC;QACH,CAAC;IACH,CAAC;SAAM,CAAC;QACN,6CAA6C;QAC7C,MAAM,QAAQ,GAAG,2BAA2B,CAAC;QAC7C,IAAI,KAAK,CAAC;QACV,OAAO,CAAC,KAAK,GAAG,QAAQ,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;YAC7C,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC;QACtB,CAAC;IACH,CAAC;IAED,OAAO,EAAE,IAAI,EAAE,gBAAgB,EAAE,CAAC;AACpC,CAAC;AAED,SAAS,wBAAwB,CAAC,GAAiB;IACjD,MAAM,YAAY,GAAG,GAAG,CAAC,eAAe,CAAC,GAAG,CAAC,iBAAiB,CAAC,CAAC;IAChE,MAAM,UAAU,GAAG,CAAC,YAAY,EAAE,OAAO,EAAE,eAAe,IAAI,EAAE,CAAqB,CAAC;IAEtF,OAAO,4BAA4B,CAAC,UAAU,CAAC,CAAC;AAClD,CAAC;AAED,SAAS,4BAA4B,CAAC,KAAuB;IAC3D,MAAM,IAAI,GAAG,IAAI,GAAG,EAAU,CAAC;IAC/B,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QACzB,MAAM,KAAK,GAAG,oBAAoB,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;QACjD,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;YACzB,IAAI,IAAI,CAAC,GAAG,CAAC,UAAU,CAAC,SAAS,CAAC,IAAI,IAAI,CAAC,GAAG,CAAC,UAAU,CAAC,UAAU,CAAC,EAAE,CAAC;gBACtE,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;YACrB,CAAC;QACH,CAAC;IACH,CAAC;IACD,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AAC1B,CAAC;AAED;;;GAGG;AACH,KAAK,UAAU,gBAAgB,CAAC,GAAiB;IAC/C,MAAM,UAAU,GAAG,IAAI,GAAG,EAAU,CAAC;IACrC,UAAU,CAAC,GAAG,CAAC,GAAG,GAAG,CAAC,OAAO,WAAW,CAAC,CAAC;IAC1C,UAAU,CAAC,GAAG,CAAC,GAAG,GAAG,CAAC,MAAM,WAAW,CAAC,CAAC;IACzC,UAAU,CAAC,GAAG,CAAC,GAAG,GAAG,CAAC,MAAM,gBAAgB,CAAC,CAAC;IAE9C,MAAM,UAAU,GAAqB,EAAE,CAAC;IACxC,KAAK,MAAM,GAAG,IAAI,UAAU,EAAE,CAAC;QAC7B,IAAI,CAAC;YACH,MAAM,QAAQ,GAAG,MAAM,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;YAC3C,IAAI,CAAC,QAAQ,CAAC,EAAE;gBAAE,SAAS;YAC3B,MAAM,WAAW,GAAG,QAAQ,CAAC,OAAO,CAAC,GAAG,CAAC,cAAc,CAAC,IAAI,EAAE,CAAC;YAC/D,IAAI,CAAC,WAAW,CAAC,QAAQ,CAAC,OAAO,CAAC;gBAAE,SAAS;YAC7C,MAAM,OAAO,GAAG,MAAM,QAAQ,CAAC,IAAI,EAAE,CAAC;YACtC,MAAM,OAAO,GAAG,OAAO,CAAC,SAAS,EAAE,CAAC,WAAW,EAAE,CAAC;YAClD,IAAI,OAAO,CAAC,UAAU,CAAC,IAAI,CAAC,IAAI,OAAO,CAAC,UAAU,CAAC,OAAO,CAAC;gBAAE,SAAS;YACtE,IAAI,OAAO,CAAC,IAAI,EAAE,CAAC,MAAM,KAAK,CAAC;gBAAE,SAAS;YAC1C,UAAU,CAAC,IAAI,CAAC;gBACd,GAAG;gBACH,OAAO;gBACP,MAAM,EAAE,QAAQ,CAAC,MAAM;gBACvB,UAAU,EAAE,QAAQ,CAAC,UAAU;aAChC,CAAC,CAAC;QACL,CAAC;QAAC,MAAM,CAAC;YACP,sBAAsB;QACxB,CAAC;IACH,CAAC;IAED,OAAO,4BAA4B,CAAC,UAAU,CAAC,CAAC;AAClD,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,sBAAsB,CAAC,SAAiB;IACtD,MAAM,IAAI,GAAa,EAAE,CAAC;IAC1B,KAAK,MAAM,IAAI,IAAI,SAAS,CAAC,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC;QACzC,MAAM,KAAK,GAAG,uBAAuB,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACjD,IAAI,KAAK,EAAE,CAAC;YACV,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC;QACtB,CAAC;IACH,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED;;GAEG;AACH,KAAK,UAAU,mBAAmB,CAAC,GAAiB;IAClD,wCAAwC;IACxC,IAAI,CAAC;QACH,MAAM,cAAc,GAAG,MAAM,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,GAAG,CAAC,MAAM,aAAa,CAAC,CAAC;QACxE,IAAI,cAAc,CAAC,EAAE,EAAE,CAAC;YACtB,MAAM,IAAI,GAAG,MAAM,cAAc,CAAC,IAAI,EAAE,CAAC;YACzC,MAAM,UAAU,GAAG,sBAAsB,CAAC,IAAI,CAAC,CAAC;YAChD,IAAI,UAAU,CAAC,MAAM,GAAG,CAAC;gBAAE,OAAO,UAAU,CAAC;QAC/C,CAAC;IACH,CAAC;IAAC,MAAM,CAAC;QACP,wCAAwC;IAC1C,CAAC;IAED,0BAA0B;IAC1B,OAAO,CAAC,GAAG,GAAG,CAAC,MAAM,cAAc,CAAC,CAAC;AACvC,CAAC;AAOD,SAAS,SAAS,CAAC,GAAW;IAC5B,OAAO,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;AACjC,CAAC;AAED,KAAK,UAAU,YAAY,CACzB,GAAiB,EACjB,UAAkB,EAClB,QAAkB;IAElB,IAAI,SAAS,CAAC,UAAU,CAAC,EAAE,CAAC;QAC1B,QAAQ,CAAC,IAAI,CAAC,4CAA4C,UAAU,EAAE,CAAC,CAAC;QACxE,OAAO,EAAE,IAAI,EAAE,EAAE,EAAE,gBAAgB,EAAE,EAAE,EAAE,CAAC;IAC5C,CAAC;IAED,IAAI,CAAC;QACH,MAAM,QAAQ,GAAG,MAAM,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,UAAU,CAAC,CAAC;QAClD,IAAI,CAAC,QAAQ,CAAC,EAAE;YAAE,OAAO,EAAE,IAAI,EAAE,EAAE,EAAE,gBAAgB,EAAE,EAAE,EAAE,CAAC;QAC5D,MAAM,GAAG,GAAG,MAAM,QAAQ,CAAC,IAAI,EAAE,CAAC;QAClC,OAAO,gBAAgB,CAAC,GAAG,CAAC,CAAC;IAC/B,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,EAAE,IAAI,EAAE,EAAE,EAAE,gBAAgB,EAAE,EAAE,EAAE,CAAC;IAC5C,CAAC;AACH,CAAC;AAED,KAAK,UAAU,kBAAkB,CAAC,GAAiB,EAAE,QAAkB;IACrE,MAAM,WAAW,GAAG,MAAM,mBAAmB,CAAC,GAAG,CAAC,CAAC;IACnD,MAAM,IAAI,GAAa,EAAE,CAAC;IAE1B,KAAK,MAAM,UAAU,IAAI,WAAW,EAAE,CAAC;QACrC,IAAI,IAAI,CAAC,MAAM,IAAI,gBAAgB;YAAE,MAAM;QAE3C,MAAM,MAAM,GAAG,MAAM,YAAY,CAAC,GAAG,EAAE,UAAU,EAAE,QAAQ,CAAC,CAAC;QAE7D,4CAA4C;QAC5C,KAAK,MAAM,GAAG,IAAI,MAAM,CAAC,IAAI,EAAE,CAAC;YAC9B,IAAI,IAAI,CAAC,MAAM,IAAI,gBAAgB;gBAAE,MAAM;YAC3C,IAAI,CAAC;gBACH,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC;gBACvB,IAAI,CAAC,CAAC,MAAM,KAAK,GAAG,CAAC,MAAM,EAAE,CAAC;oBAC5B,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;gBACjB,CAAC;YACH,CAAC;YAAC,MAAM,CAAC;gBACP,sBAAsB;YACxB,CAAC;QACH,CAAC;QAED,oCAAoC;QACpC,IAAI,MAAM,CAAC,gBAAgB,CAAC,MAAM,GAAG,CAAC,IAAI,IAAI,CAAC,MAAM,GAAG,gBAAgB,EAAE,CAAC;YACzE,KAAK,MAAM,aAAa,IAAI,MAAM,CAAC,gBAAgB,EAAE,CAAC;gBACpD,IAAI,IAAI,CAAC,MAAM,IAAI,gBAAgB;oBAAE,MAAM;gBAE3C,MAAM,SAAS,GAAG,MAAM,YAAY,CAAC,GAAG,EAAE,aAAa,EAAE,QAAQ,CAAC,CAAC;gBAEnE,KAAK,MAAM,GAAG,IAAI,SAAS,CAAC,IAAI,EAAE,CAAC;oBACjC,IAAI,IAAI,CAAC,MAAM,IAAI,gBAAgB;wBAAE,MAAM;oBAC3C,IAAI,CAAC;wBACH,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC;wBACvB,IAAI,CAAC,CAAC,MAAM,KAAK,GAAG,CAAC,MAAM,EAAE,CAAC;4BAC5B,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;wBACjB,CAAC;oBACH,CAAC;oBAAC,MAAM,CAAC;wBACP,sBAAsB;oBACxB,CAAC;gBACH,CAAC;YACH,CAAC;QACH,CAAC;IACH,CAAC;IAED,OAAO,IAAI,CAAC;AACd,CAAC;AAED;;;;;;;GAOG;AACH,MAAM,CAAC,KAAK,UAAU,WAAW,CAAC,GAAiB;IACjD,MAAM,QAAQ,GAAa,EAAE,CAAC;IAE9B,qEAAqE;IACrE,MAAM,UAAU,GAAG,wBAAwB,CAAC,GAAG,CAAC,CAAC;IACjD,IAAI,UAAU,CAAC,MAAM,GAAG,CAAC;QAAE,OAAO,EAAE,IAAI,EAAE,UAAU,EAAE,QAAQ,EAAE,CAAC;IAEjE,kFAAkF;IAClF,IAAI,CAAC,GAAG,CAAC,eAAe,CAAC,GAAG,CAAC,iBAAiB,CAAC,EAAE,CAAC;QAChD,MAAM,WAAW,GAAG,MAAM,gBAAgB,CAAC,GAAG,CAAC,CAAC;QAChD,IAAI,WAAW,CAAC,MAAM,GAAG,CAAC;YAAE,OAAO,EAAE,IAAI,EAAE,WAAW,EAAE,QAAQ,EAAE,CAAC;IACrE,CAAC;IAED,iBAAiB;IACjB,MAAM,WAAW,GAAG,MAAM,kBAAkB,CAAC,GAAG,EAAE,QAAQ,CAAC,CAAC;IAC5D,IAAI,WAAW,CAAC,MAAM,GAAG,CAAC;QAAE,OAAO,EAAE,IAAI,EAAE,WAAW,EAAE,QAAQ,EAAE,CAAC;IAEnE,cAAc;IACd,OAAO,EAAE,IAAI,EAAE,CAAC,GAAG,CAAC,OAAO,CAAC,EAAE,QAAQ,EAAE,CAAC;AAC3C,CAAC;AASD;;;GAGG;AACH,MAAM,CAAC,KAAK,UAAU,sBAAsB,CAAC,GAAiB;IAC5D,MAAM,SAAS,GAAG,MAAM,WAAW,CAAC,GAAG,CAAC,CAAC;IACzC,IAAI,IAAI,GAAG,SAAS,CAAC,IAAI,CAAC;IAC1B,MAAM,UAAU,GAAG,IAAI,CAAC,MAAM,CAAC;IAE/B,MAAM,OAAO,GAAG,UAAU,GAAG,GAAG,CAAC,OAAO,CAAC,cAAc,CAAC;IACxD,IAAI,OAAO,EAAE,CAAC;QACZ,KAAK,IAAI,CAAC,GAAG,IAAI,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;YACzC,MAAM,CAAC,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;YAC9C,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;QAC1C,CAAC;QACD,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,GAAG,CAAC,OAAO,CAAC,cAAc,CAAC,CAAC;IACnD,CAAC;IAED,OAAO,EAAE,IAAI,EAAE,UAAU,EAAE,OAAO,EAAE,QAAQ,EAAE,SAAS,CAAC,QAAQ,EAAE,CAAC;AACrE,CAAC"}

package/dist/helpers/html-to-markdown.d.ts

@@ -0,0 +1,7 @@
+/**
+ * Convert HTML to markdown using Turndown with default configuration.
+ * Matches real agent behavior per the Agent-Friendly Documentation Spec:
+ * no explicit <style>/<script> stripping, default options only.
+ */
+export declare function htmlToMarkdown(html: string): string;
+//# sourceMappingURL=html-to-markdown.d.ts.map

package/dist/helpers/html-to-markdown.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"html-to-markdown.d.ts","sourceRoot":"","sources":["../../src/helpers/html-to-markdown.ts"],"names":[],"mappings":"AAEA;;;;GAIG;AACH,wBAAgB,cAAc,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,CAGnD"}

package/dist/helpers/html-to-markdown.js

@@ -0,0 +1,11 @@
+import TurndownService from 'turndown';
+/**
+ * Convert HTML to markdown using Turndown with default configuration.
+ * Matches real agent behavior per the Agent-Friendly Documentation Spec:
+ * no explicit <style>/<script> stripping, default options only.
+ */
+export function htmlToMarkdown(html) {
+    const turndown = new TurndownService();
+    return turndown.turndown(html);
+}
+//# sourceMappingURL=html-to-markdown.js.map

package/dist/helpers/html-to-markdown.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"html-to-markdown.js","sourceRoot":"","sources":["../../src/helpers/html-to-markdown.ts"],"names":[],"mappings":"AAAA,OAAO,eAAe,MAAM,UAAU,CAAC;AAEvC;;;;GAIG;AACH,MAAM,UAAU,cAAc,CAAC,IAAY;IACzC,MAAM,QAAQ,GAAG,IAAI,eAAe,EAAE,CAAC;IACvC,OAAO,QAAQ,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC;AACjC,CAAC"}

package/dist/helpers/index.d.ts

@@ -1,3 +1,8 @@
 export { loadConfig } from './config.js';
 export { describeAgentDocs, describeAgentDocsPerCheck } from './vitest-runner.js';
+export { looksLikeMarkdown, looksLikeHtml } from './detect-markdown.js';
+export { getPageUrls, discoverAndSamplePages, parseSitemapUrls, parseSitemapDirectives, } from './get-page-urls.js';
+export type { PageUrlResult, SampledPages } from './get-page-urls.js';
+export { toMdUrls } from './to-md-urls.js';
+export { htmlToMarkdown } from './html-to-markdown.js';
 //# sourceMappingURL=index.d.ts.map

package/dist/helpers/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/helpers/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,UAAU,EAAE,MAAM,aAAa,CAAC;AACzC,OAAO,EAAE,iBAAiB,EAAE,yBAAyB,EAAE,MAAM,oBAAoB,CAAC"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/helpers/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,UAAU,EAAE,MAAM,aAAa,CAAC;AACzC,OAAO,EAAE,iBAAiB,EAAE,yBAAyB,EAAE,MAAM,oBAAoB,CAAC;AAClF,OAAO,EAAE,iBAAiB,EAAE,aAAa,EAAE,MAAM,sBAAsB,CAAC;AACxE,OAAO,EACL,WAAW,EACX,sBAAsB,EACtB,gBAAgB,EAChB,sBAAsB,GACvB,MAAM,oBAAoB,CAAC;AAC5B,YAAY,EAAE,aAAa,EAAE,YAAY,EAAE,MAAM,oBAAoB,CAAC;AACtE,OAAO,EAAE,QAAQ,EAAE,MAAM,iBAAiB,CAAC;AAC3C,OAAO,EAAE,cAAc,EAAE,MAAM,uBAAuB,CAAC"}

package/dist/helpers/index.js

@@ -1,3 +1,7 @@
 export { loadConfig } from './config.js';
 export { describeAgentDocs, describeAgentDocsPerCheck } from './vitest-runner.js';
+export { looksLikeMarkdown, looksLikeHtml } from './detect-markdown.js';
+export { getPageUrls, discoverAndSamplePages, parseSitemapUrls, parseSitemapDirectives, } from './get-page-urls.js';
+export { toMdUrls } from './to-md-urls.js';
+export { htmlToMarkdown } from './html-to-markdown.js';
 //# sourceMappingURL=index.js.map

package/dist/helpers/index.js.map

@@ -1 +1 @@
-{"version":3,"file":"index.js","sourceRoot":"","sources":["../../src/helpers/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,UAAU,EAAE,MAAM,aAAa,CAAC;AACzC,OAAO,EAAE,iBAAiB,EAAE,yBAAyB,EAAE,MAAM,oBAAoB,CAAC"}
+{"version":3,"file":"index.js","sourceRoot":"","sources":["../../src/helpers/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,UAAU,EAAE,MAAM,aAAa,CAAC;AACzC,OAAO,EAAE,iBAAiB,EAAE,yBAAyB,EAAE,MAAM,oBAAoB,CAAC;AAClF,OAAO,EAAE,iBAAiB,EAAE,aAAa,EAAE,MAAM,sBAAsB,CAAC;AACxE,OAAO,EACL,WAAW,EACX,sBAAsB,EACtB,gBAAgB,EAChB,sBAAsB,GACvB,MAAM,oBAAoB,CAAC;AAE5B,OAAO,EAAE,QAAQ,EAAE,MAAM,iBAAiB,CAAC;AAC3C,OAAO,EAAE,cAAc,EAAE,MAAM,uBAAuB,CAAC"}

package/dist/helpers/to-md-urls.d.ts

@@ -0,0 +1,7 @@
+/**
+ * Generate candidate .md URLs for a page URL.
+ * If the URL already ends in .md, return it as-is.
+ * Otherwise try both `/docs/guide.md` and `/docs/guide/index.md`.
+ */
+export declare function toMdUrls(url: string): string[];
+//# sourceMappingURL=to-md-urls.d.ts.map

package/dist/helpers/to-md-urls.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"to-md-urls.d.ts","sourceRoot":"","sources":["../../src/helpers/to-md-urls.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AACH,wBAAgB,QAAQ,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,EAAE,CAsB9C"}

package/dist/helpers/to-md-urls.js

@@ -0,0 +1,24 @@
+/**
+ * Generate candidate .md URLs for a page URL.
+ * If the URL already ends in .md, return it as-is.
+ * Otherwise try both `/docs/guide.md` and `/docs/guide/index.md`.
+ */
+export function toMdUrls(url) {
+    const parsed = new URL(url);
+    // URL already points to a .md file — use it directly
+    if (parsed.pathname.endsWith('.md')) {
+        return [url];
+    }
+    const pathname = parsed.pathname.replace(/\/$/, '') || '';
+    const candidates = [];
+    // /docs/guide.md (strip .html extension if present)
+    const directMd = new URL(parsed.toString());
+    directMd.pathname = pathname.replace(/\.html?$/, '') + '.md';
+    candidates.push(directMd.toString());
+    // /docs/guide/index.md
+    const indexMd = new URL(parsed.toString());
+    indexMd.pathname = pathname + '/index.md';
+    candidates.push(indexMd.toString());
+    return candidates;
+}
+//# sourceMappingURL=to-md-urls.js.map

package/dist/helpers/to-md-urls.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"to-md-urls.js","sourceRoot":"","sources":["../../src/helpers/to-md-urls.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AACH,MAAM,UAAU,QAAQ,CAAC,GAAW;IAClC,MAAM,MAAM,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC;IAE5B,qDAAqD;IACrD,IAAI,MAAM,CAAC,QAAQ,CAAC,QAAQ,CAAC,KAAK,CAAC,EAAE,CAAC;QACpC,OAAO,CAAC,GAAG,CAAC,CAAC;IACf,CAAC;IAED,MAAM,QAAQ,GAAG,MAAM,CAAC,QAAQ,CAAC,OAAO,CAAC,KAAK,EAAE,EAAE,CAAC,IAAI,EAAE,CAAC;IAC1D,MAAM,UAAU,GAAa,EAAE,CAAC;IAEhC,oDAAoD;IACpD,MAAM,QAAQ,GAAG,IAAI,GAAG,CAAC,MAAM,CAAC,QAAQ,EAAE,CAAC,CAAC;IAC5C,QAAQ,CAAC,QAAQ,GAAG,QAAQ,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,GAAG,KAAK,CAAC;IAC7D,UAAU,CAAC,IAAI,CAAC,QAAQ,CAAC,QAAQ,EAAE,CAAC,CAAC;IAErC,uBAAuB;IACvB,MAAM,OAAO,GAAG,IAAI,GAAG,CAAC,MAAM,CAAC,QAAQ,EAAE,CAAC,CAAC;IAC3C,OAAO,CAAC,QAAQ,GAAG,QAAQ,GAAG,WAAW,CAAC;IAC1C,UAAU,CAAC,IAAI,CAAC,OAAO,CAAC,QAAQ,EAAE,CAAC,CAAC;IAEpC,OAAO,UAAU,CAAC;AACpB,CAAC"}

package/dist/runner.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"runner.d.ts","sourceRoot":"","sources":["../src/runner.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,YAAY,EAAe,aAAa,EAAE,YAAY,EAAE,MAAM,YAAY,CAAC;AAmCzF,wBAAgB,aAAa,CAAC,OAAO,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,OAAO,CAAC,aAAa,CAAC,GAAG,YAAY,CAe7F;AAED,wBAAsB,SAAS,CAC7B,OAAO,EAAE,MAAM,EACf,OAAO,CAAC,EAAE,OAAO,CAAC,aAAa,CAAC,GAC/B,OAAO,CAAC,YAAY,CAAC,CA0DvB"}
+{"version":3,"file":"runner.d.ts","sourceRoot":"","sources":["../src/runner.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,YAAY,EAAe,aAAa,EAAE,YAAY,EAAE,MAAM,YAAY,CAAC;AAmCzF,wBAAgB,aAAa,CAAC,OAAO,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,OAAO,CAAC,aAAa,CAAC,GAAG,YAAY,CAgB7F;AAED,wBAAsB,SAAS,CAC7B,OAAO,EAAE,MAAM,EACf,OAAO,CAAC,EAAE,OAAO,CAAC,aAAa,CAAC,GAC/B,OAAO,CAAC,YAAY,CAAC,CAkEvB"}

package/dist/runner.js

@@ -18,10 +18,10 @@ function normalizeDeps(deps) {
 function checkDependenciesMet(deps, previousResults) {
     const normalized = normalizeDeps(deps);
     for (const orGroup of normalized) {
-        // At least one check in the OR-group must have passed
+        // At least one check in the OR-group must have passed (or warned)
         const anyPassed = orGroup.some((id) => {
             const result = previousResults.get(id);
-            return result?.status === 'pass';
+            return result?.status === 'pass' || result?.status === 'warn';
         });
         if (!anyPassed)
             return false;
@@ -41,6 +41,7 @@ export function createContext(baseUrl, options) {
             maxConcurrency: merged.maxConcurrency,
         }),
         options: merged,
+        pageCache: new Map(),
     };
 }
 export async function runChecks(baseUrl, options) {
@@ -53,18 +54,24 @@ export async function runChecks(baseUrl, options) {
         if (checkIds && checkIds.length > 0 && !checkIds.includes(check.id)) {
             continue;
         }
-        // Check dependencies
-        if (check.dependsOn.length > 0 && !checkDependenciesMet(check.dependsOn, ctx.previousResults)) {
-            const result = {
-                id: check.id,
-                category: check.category,
-                status: 'skip',
-                message: 'Skipped: dependency check did not pass',
-                dependsOn: normalizeDeps(check.dependsOn).flat(),
-            };
-            results.push(result);
-            ctx.previousResults.set(check.id, result);
-            continue;
+        // Check dependencies — only skip if at least one dependency actually ran and none passed.
+        // If no dependencies ran at all (e.g. filtered out via --checks), let the check handle
+        // standalone mode itself.
+        if (check.dependsOn.length > 0) {
+            const normalized = normalizeDeps(check.dependsOn);
+            const anyDepRan = normalized.some((orGroup) => orGroup.some((id) => ctx.previousResults.has(id)));
+            if (anyDepRan && !checkDependenciesMet(check.dependsOn, ctx.previousResults)) {
+                const result = {
+                    id: check.id,
+                    category: check.category,
+                    status: 'skip',
+                    message: 'Skipped: dependency check did not pass',
+                    dependsOn: normalized.flat(),
+                };
+                results.push(result);
+                ctx.previousResults.set(check.id, result);
+                continue;
+            }
         }
         try {
             const result = await check.run(ctx);

package/dist/runner.js.map

@@ -1 +1 @@
-{"version":3,"file":"runner.js","sourceRoot":"","sources":["../src/runner.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,eAAe,EAAE,MAAM,gBAAgB,CAAC;AACjD,OAAO,EAAE,gBAAgB,EAAE,MAAM,WAAW,CAAC;AAC7C,OAAO,EAAE,eAAe,EAAE,MAAM,sBAAsB,CAAC;AAEvD;;;;GAIG;AACH,SAAS,aAAa,CAAC,IAA2B;IAChD,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IACjC,IAAI,OAAO,IAAI,CAAC,CAAC,CAAC,KAAK,QAAQ,EAAE,CAAC;QAChC,sDAAsD;QACtD,OAAQ,IAAiB,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC5C,CAAC;IACD,OAAO,IAAkB,CAAC;AAC5B,CAAC;AAED,SAAS,oBAAoB,CAC3B,IAA2B,EAC3B,eAAyC;IAEzC,MAAM,UAAU,GAAG,aAAa,CAAC,IAAI,CAAC,CAAC;IACvC,KAAK,MAAM,OAAO,IAAI,UAAU,EAAE,CAAC;QACjC,sDAAsD;QACtD,MAAM,SAAS,GAAG,OAAO,CAAC,IAAI,CAAC,CAAC,EAAE,EAAE,EAAE;YACpC,MAAM,MAAM,GAAG,eAAe,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;YACvC,OAAO,MAAM,EAAE,MAAM,KAAK,MAAM,CAAC;QACnC,CAAC,CAAC,CAAC;QACH,IAAI,CAAC,SAAS;YAAE,OAAO,KAAK,CAAC;IAC/B,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED,MAAM,UAAU,aAAa,CAAC,OAAe,EAAE,OAAgC;IAC7E,MAAM,MAAM,GAAG,EAAE,GAAG,eAAe,EAAE,GAAG,OAAO,EAAE,CAAC;IAClD,MAAM,GAAG,GAAG,IAAI,GAAG,CAAC,OAAO,CAAC,CAAC;IAE7B,OAAO;QACL,OAAO,EAAE,OAAO,CAAC,OAAO,CAAC,KAAK,EAAE,EAAE,CAAC;QACnC,MAAM,EAAE,GAAG,CAAC,MAAM;QAClB,eAAe,EAAE,IAAI,GAAG,EAAE;QAC1B,IAAI,EAAE,gBAAgB,CAAC;YACrB,YAAY,EAAE,MAAM,CAAC,YAAY;YACjC,cAAc,EAAE,MAAM,CAAC,cAAc;YACrC,cAAc,EAAE,MAAM,CAAC,cAAc;SACtC,CAAC;QACF,OAAO,EAAE,MAAM;KAChB,CAAC;AACJ,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,SAAS,CAC7B,OAAe,EACf,OAAgC;IAEhC,MAAM,GAAG,GAAG,aAAa,CAAC,OAAO,EAAE,OAAO,CAAC,CAAC;IAC5C,MAAM,SAAS,GAAG,eAAe,EAAE,CAAC;IACpC,MAAM,QAAQ,GAAG,OAAO,EAAE,QAAQ,CAAC;IAEnC,MAAM,OAAO,GAAkB,EAAE,CAAC;IAElC,KAAK,MAAM,KAAK,IAAI,SAAS,EAAE,CAAC;QAC9B,4CAA4C;QAC5C,IAAI,QAAQ,IAAI,QAAQ,CAAC,MAAM,GAAG,CAAC,IAAI,CAAC,QAAQ,CAAC,QAAQ,CAAC,KAAK,CAAC,EAAE,CAAC,EAAE,CAAC;YACpE,SAAS;QACX,CAAC;QAED,qBAAqB;QACrB,IAAI,KAAK,CAAC,SAAS,CAAC,MAAM,GAAG,CAAC,IAAI,CAAC,oBAAoB,CAAC,KAAK,CAAC,SAAS,EAAE,GAAG,CAAC,eAAe,CAAC,EAAE,CAAC;YAC9F,MAAM,MAAM,GAAgB;gBAC1B,EAAE,EAAE,KAAK,CAAC,EAAE;gBACZ,QAAQ,EAAE,KAAK,CAAC,QAAQ;gBACxB,MAAM,EAAE,MAAM;gBACd,OAAO,EAAE,wCAAwC;gBACjD,SAAS,EAAE,aAAa,CAAC,KAAK,CAAC,SAAS,CAAC,CAAC,IAAI,EAAE;aACjD,CAAC;YACF,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;YACrB,GAAG,CAAC,eAAe,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,EAAE,MAAM,CAAC,CAAC;YAC1C,SAAS;QACX,CAAC;QAED,IAAI,CAAC;YACH,MAAM,MAAM,GAAG,MAAM,KAAK,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;YACpC,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;YACrB,GAAG,CAAC,eAAe,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,EAAE,MAAM,CAAC,CAAC;QAC5C,CAAC;QAAC,OAAO,GAAG,EAAE,CAAC;YACb,MAAM,MAAM,GAAgB;gBAC1B,EAAE,EAAE,KAAK,CAAC,EAAE;gBACZ,QAAQ,EAAE,KAAK,CAAC,QAAQ;gBACxB,MAAM,EAAE,OAAO;gBACf,OAAO,EAAE,gBAAgB,GAAG,YAAY,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,EAAE;aAC5E,CAAC;YACF,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;YACrB,GAAG,CAAC,eAAe,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,EAAE,MAAM,CAAC,CAAC;QAC5C,CAAC;IACH,CAAC;IAED,MAAM,OAAO,GAAG;QACd,KAAK,EAAE,OAAO,CAAC,MAAM;QACrB,IAAI,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,KAAK,MAAM,CAAC,CAAC,MAAM;QACvD,IAAI,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,KAAK,MAAM,CAAC,CAAC,MAAM;QACvD,IAAI,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,KAAK,MAAM,CAAC,CAAC,MAAM;QACvD,IAAI,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,KAAK,MAAM,CAAC,CAAC,MAAM;QACvD,KAAK,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,KAAK,OAAO,CAAC,CAAC,MAAM;KAC1D,CAAC;IAEF,OAAO;QACL,GAAG,EAAE,OAAO;QACZ,SAAS,EAAE,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE;QACnC,OAAO;QACP,OAAO;KACR,CAAC;AACJ,CAAC"}
+{"version":3,"file":"runner.js","sourceRoot":"","sources":["../src/runner.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,eAAe,EAAE,MAAM,gBAAgB,CAAC;AACjD,OAAO,EAAE,gBAAgB,EAAE,MAAM,WAAW,CAAC;AAC7C,OAAO,EAAE,eAAe,EAAE,MAAM,sBAAsB,CAAC;AAEvD;;;;GAIG;AACH,SAAS,aAAa,CAAC,IAA2B;IAChD,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IACjC,IAAI,OAAO,IAAI,CAAC,CAAC,CAAC,KAAK,QAAQ,EAAE,CAAC;QAChC,sDAAsD;QACtD,OAAQ,IAAiB,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC5C,CAAC;IACD,OAAO,IAAkB,CAAC;AAC5B,CAAC;AAED,SAAS,oBAAoB,CAC3B,IAA2B,EAC3B,eAAyC;IAEzC,MAAM,UAAU,GAAG,aAAa,CAAC,IAAI,CAAC,CAAC;IACvC,KAAK,MAAM,OAAO,IAAI,UAAU,EAAE,CAAC;QACjC,kEAAkE;QAClE,MAAM,SAAS,GAAG,OAAO,CAAC,IAAI,CAAC,CAAC,EAAE,EAAE,EAAE;YACpC,MAAM,MAAM,GAAG,eAAe,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;YACvC,OAAO,MAAM,EAAE,MAAM,KAAK,MAAM,IAAI,MAAM,EAAE,MAAM,KAAK,MAAM,CAAC;QAChE,CAAC,CAAC,CAAC;QACH,IAAI,CAAC,SAAS;YAAE,OAAO,KAAK,CAAC;IAC/B,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED,MAAM,UAAU,aAAa,CAAC,OAAe,EAAE,OAAgC;IAC7E,MAAM,MAAM,GAAG,EAAE,GAAG,eAAe,EAAE,GAAG,OAAO,EAAE,CAAC;IAClD,MAAM,GAAG,GAAG,IAAI,GAAG,CAAC,OAAO,CAAC,CAAC;IAE7B,OAAO;QACL,OAAO,EAAE,OAAO,CAAC,OAAO,CAAC,KAAK,EAAE,EAAE,CAAC;QACnC,MAAM,EAAE,GAAG,CAAC,MAAM;QAClB,eAAe,EAAE,IAAI,GAAG,EAAE;QAC1B,IAAI,EAAE,gBAAgB,CAAC;YACrB,YAAY,EAAE,MAAM,CAAC,YAAY;YACjC,cAAc,EAAE,MAAM,CAAC,cAAc;YACrC,cAAc,EAAE,MAAM,CAAC,cAAc;SACtC,CAAC;QACF,OAAO,EAAE,MAAM;QACf,SAAS,EAAE,IAAI,GAAG,EAAE;KACrB,CAAC;AACJ,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,SAAS,CAC7B,OAAe,EACf,OAAgC;IAEhC,MAAM,GAAG,GAAG,aAAa,CAAC,OAAO,EAAE,OAAO,CAAC,CAAC;IAC5C,MAAM,SAAS,GAAG,eAAe,EAAE,CAAC;IACpC,MAAM,QAAQ,GAAG,OAAO,EAAE,QAAQ,CAAC;IAEnC,MAAM,OAAO,GAAkB,EAAE,CAAC;IAElC,KAAK,MAAM,KAAK,IAAI,SAAS,EAAE,CAAC;QAC9B,4CAA4C;QAC5C,IAAI,QAAQ,IAAI,QAAQ,CAAC,MAAM,GAAG,CAAC,IAAI,CAAC,QAAQ,CAAC,QAAQ,CAAC,KAAK,CAAC,EAAE,CAAC,EAAE,CAAC;YACpE,SAAS;QACX,CAAC;QAED,0FAA0F;QAC1F,uFAAuF;QACvF,0BAA0B;QAC1B,IAAI,KAAK,CAAC,SAAS,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YAC/B,MAAM,UAAU,GAAG,aAAa,CAAC,KAAK,CAAC,SAAS,CAAC,CAAC;YAClD,MAAM,SAAS,GAAG,UAAU,CAAC,IAAI,CAAC,CAAC,OAAO,EAAE,EAAE,CAC5C,OAAO,CAAC,IAAI,CAAC,CAAC,EAAE,EAAE,EAAE,CAAC,GAAG,CAAC,eAAe,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,CAClD,CAAC;YACF,IAAI,SAAS,IAAI,CAAC,oBAAoB,CAAC,KAAK,CAAC,SAAS,EAAE,GAAG,CAAC,eAAe,CAAC,EAAE,CAAC;gBAC7E,MAAM,MAAM,GAAgB;oBAC1B,EAAE,EAAE,KAAK,CAAC,EAAE;oBACZ,QAAQ,EAAE,KAAK,CAAC,QAAQ;oBACxB,MAAM,EAAE,MAAM;oBACd,OAAO,EAAE,wCAAwC;oBACjD,SAAS,EAAE,UAAU,CAAC,IAAI,EAAE;iBAC7B,CAAC;gBACF,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;gBACrB,GAAG,CAAC,eAAe,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,EAAE,MAAM,CAAC,CAAC;gBAC1C,SAAS;YACX,CAAC;QACH,CAAC;QAED,IAAI,CAAC;YACH,MAAM,MAAM,GAAG,MAAM,KAAK,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;YACpC,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;YACrB,GAAG,CAAC,eAAe,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,EAAE,MAAM,CAAC,CAAC;QAC5C,CAAC;QAAC,OAAO,GAAG,EAAE,CAAC;YACb,MAAM,MAAM,GAAgB;gBAC1B,EAAE,EAAE,KAAK,CAAC,EAAE;gBACZ,QAAQ,EAAE,KAAK,CAAC,QAAQ;gBACxB,MAAM,EAAE,OAAO;gBACf,OAAO,EAAE,gBAAgB,GAAG,YAAY,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,EAAE;aAC5E,CAAC;YACF,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;YACrB,GAAG,CAAC,eAAe,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,EAAE,MAAM,CAAC,CAAC;QAC5C,CAAC;IACH,CAAC;IAED,MAAM,OAAO,GAAG;QACd,KAAK,EAAE,OAAO,CAAC,MAAM;QACrB,IAAI,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,KAAK,MAAM,CAAC,CAAC,MAAM;QACvD,IAAI,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,KAAK,MAAM,CAAC,CAAC,MAAM;QACvD,IAAI,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,KAAK,MAAM,CAAC,CAAC,MAAM;QACvD,IAAI,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,KAAK,MAAM,CAAC,CAAC,MAAM;QACvD,KAAK,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,MAAM,KAAK,OAAO,CAAC,CAAC,MAAM;KAC1D,CAAC;IAEF,OAAO;QACL,GAAG,EAAE,OAAO;QACZ,SAAS,EAAE,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE;QACnC,OAAO;QACP,OAAO;KACR,CAAC;AACJ,CAAC"}

package/dist/types.d.ts

@@ -7,6 +7,13 @@ export interface CheckResult {
     details?: Record<string, unknown>;
     dependsOn?: string[];
 }
+export interface CachedPage {
+    url: string;
+    markdown?: {
+        content: string;
+        source: 'md-url' | 'content-negotiation';
+    };
+}
 export interface CheckContext {
     /** The base URL being checked (as provided by the user). */
     baseUrl: string;
@@ -18,6 +25,8 @@ export interface CheckContext {
     http: HttpClient;
     /** Runtime options. */
     options: CheckOptions;
+    /** Cached page content, keyed by original page URL. */
+    pageCache: Map<string, CachedPage>;
 }
 export interface CheckOptions {
     /** Maximum concurrent HTTP requests within a single check. */
@@ -42,7 +51,7 @@ export interface CheckDefinition {
     id: string;
     category: string;
     description: string;
-    /** Check IDs that must pass before this check runs. Array of arrays for OR-groups. */
+    /** Check IDs that must pass or warn before this check runs. Array of arrays for OR-groups. */
     dependsOn: string[][] | string[];
     run: CheckFunction;
 }

package/dist/types.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../src/types.ts"],"names":[],"mappings":"AAAA,MAAM,MAAM,WAAW,GAAG,MAAM,GAAG,MAAM,GAAG,MAAM,GAAG,MAAM,GAAG,OAAO,CAAC;AAEtE,MAAM,WAAW,WAAW;IAC1B,EAAE,EAAE,MAAM,CAAC;IACX,QAAQ,EAAE,MAAM,CAAC;IACjB,MAAM,EAAE,WAAW,CAAC;IACpB,OAAO,EAAE,MAAM,CAAC;IAChB,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;IAClC,SAAS,CAAC,EAAE,MAAM,EAAE,CAAC;CACtB;AAED,MAAM,WAAW,YAAY;IAC3B,4DAA4D;IAC5D,OAAO,EAAE,MAAM,CAAC;IAChB,uDAAuD;IACvD,MAAM,EAAE,MAAM,CAAC;IACf,6DAA6D;IAC7D,eAAe,EAAE,GAAG,CAAC,MAAM,EAAE,WAAW,CAAC,CAAC;IAC1C,sCAAsC;IACtC,IAAI,EAAE,UAAU,CAAC;IACjB,uBAAuB;IACvB,OAAO,EAAE,YAAY,CAAC;CACvB;AAED,MAAM,WAAW,YAAY;IAC3B,8DAA8D;IAC9D,cAAc,EAAE,MAAM,CAAC;IACvB,yCAAyC;IACzC,YAAY,EAAE,MAAM,CAAC;IACrB,kDAAkD;IAClD,cAAc,EAAE,MAAM,CAAC;IACvB,iEAAiE;IACjE,cAAc,EAAE,MAAM,CAAC;IACvB,uBAAuB;IACvB,UAAU,EAAE,cAAc,CAAC;CAC5B;AAED,MAAM,WAAW,cAAc;IAC7B,8DAA8D;IAC9D,IAAI,EAAE,MAAM,CAAC;IACb,8DAA8D;IAC9D,IAAI,EAAE,MAAM,CAAC;CACd;AAED,MAAM,MAAM,aAAa,GAAG,CAAC,GAAG,EAAE,YAAY,KAAK,OAAO,CAAC,WAAW,CAAC,CAAC;AAExE,MAAM,WAAW,eAAe;IAC9B,EAAE,EAAE,MAAM,CAAC;IACX,QAAQ,EAAE,MAAM,CAAC;IACjB,WAAW,EAAE,MAAM,CAAC;IACpB,sFAAsF;IACtF,SAAS,EAAE,MAAM,EAAE,EAAE,GAAG,MAAM,EAAE,CAAC;IACjC,GAAG,EAAE,aAAa,CAAC;CACpB;AAED,MAAM,WAAW,UAAU;IACzB,KAAK,CAAC,GAAG,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,kBAAkB,GAAG,OAAO,CAAC,YAAY,CAAC,CAAC;CACzE;AAED,MAAM,WAAW,kBAAkB;IACjC,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IACjC,QAAQ,CAAC,EAAE,QAAQ,GAAG,QAAQ,CAAC;IAC/B,MAAM,CAAC,EAAE,WAAW,CAAC;CACtB;AAED,MAAM,WAAW,YAAY;IAC3B,EAAE,EAAE,OAAO,CAAC;IACZ,MAAM,EAAE,MAAM,CAAC;IACf,UAAU,EAAE,MAAM,CAAC;IACnB,OAAO,EAAE,OAAO,CAAC;IACjB,GAAG,EAAE,MAAM,CAAC;IACZ,UAAU,EAAE,OAAO,CAAC;IACpB,IAAI,IAAI,OAAO,CAAC,MAAM,CAAC,CAAC;CACzB;AAED,MAAM,WAAW,cAAc;IAC7B,GAAG,EAAE,MAAM,CAAC;IACZ,OAAO,EAAE,MAAM,CAAC;IAChB,MAAM,EAAE,MAAM,CAAC;IACf,UAAU,EAAE,OAAO,CAAC;IACpB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,iBAAiB,CAAC,EAAE,OAAO,CAAC;CAC7B;AAED,MAAM,WAAW,aAAc,SAAQ,YAAY;IACjD,6DAA6D;IAC7D,QAAQ,CAAC,EAAE,MAAM,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,YAAY;IAC3B,GAAG,EAAE,MAAM,CAAC;IACZ,SAAS,EAAE,MAAM,CAAC;IAClB,OAAO,EAAE,WAAW,EAAE,CAAC;IACvB,OAAO,EAAE;QACP,KAAK,EAAE,MAAM,CAAC;QACd,IAAI,EAAE,MAAM,CAAC;QACb,IAAI,EAAE,MAAM,CAAC;QACb,IAAI,EAAE,MAAM,CAAC;QACb,IAAI,EAAE,MAAM,CAAC;QACb,KAAK,EAAE,MAAM,CAAC;KACf,CAAC;CACH;AAED,MAAM,WAAW,eAAe;IAC9B,GAAG,EAAE,MAAM,CAAC;IACZ,MAAM,CAAC,EAAE,MAAM,EAAE,CAAC;IAClB,OAAO,CAAC,EAAE,OAAO,CAAC,YAAY,CAAC,CAAC;CACjC"}
+{"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../src/types.ts"],"names":[],"mappings":"AAAA,MAAM,MAAM,WAAW,GAAG,MAAM,GAAG,MAAM,GAAG,MAAM,GAAG,MAAM,GAAG,OAAO,CAAC;AAEtE,MAAM,WAAW,WAAW;IAC1B,EAAE,EAAE,MAAM,CAAC;IACX,QAAQ,EAAE,MAAM,CAAC;IACjB,MAAM,EAAE,WAAW,CAAC;IACpB,OAAO,EAAE,MAAM,CAAC;IAChB,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;IAClC,SAAS,CAAC,EAAE,MAAM,EAAE,CAAC;CACtB;AAED,MAAM,WAAW,UAAU;IACzB,GAAG,EAAE,MAAM,CAAC;IACZ,QAAQ,CAAC,EAAE;QACT,OAAO,EAAE,MAAM,CAAC;QAChB,MAAM,EAAE,QAAQ,GAAG,qBAAqB,CAAC;KAC1C,CAAC;CACH;AAED,MAAM,WAAW,YAAY;IAC3B,4DAA4D;IAC5D,OAAO,EAAE,MAAM,CAAC;IAChB,uDAAuD;IACvD,MAAM,EAAE,MAAM,CAAC;IACf,6DAA6D;IAC7D,eAAe,EAAE,GAAG,CAAC,MAAM,EAAE,WAAW,CAAC,CAAC;IAC1C,sCAAsC;IACtC,IAAI,EAAE,UAAU,CAAC;IACjB,uBAAuB;IACvB,OAAO,EAAE,YAAY,CAAC;IACtB,uDAAuD;IACvD,SAAS,EAAE,GAAG,CAAC,MAAM,EAAE,UAAU,CAAC,CAAC;CACpC;AAED,MAAM,WAAW,YAAY;IAC3B,8DAA8D;IAC9D,cAAc,EAAE,MAAM,CAAC;IACvB,yCAAyC;IACzC,YAAY,EAAE,MAAM,CAAC;IACrB,kDAAkD;IAClD,cAAc,EAAE,MAAM,CAAC;IACvB,iEAAiE;IACjE,cAAc,EAAE,MAAM,CAAC;IACvB,uBAAuB;IACvB,UAAU,EAAE,cAAc,CAAC;CAC5B;AAED,MAAM,WAAW,cAAc;IAC7B,8DAA8D;IAC9D,IAAI,EAAE,MAAM,CAAC;IACb,8DAA8D;IAC9D,IAAI,EAAE,MAAM,CAAC;CACd;AAED,MAAM,MAAM,aAAa,GAAG,CAAC,GAAG,EAAE,YAAY,KAAK,OAAO,CAAC,WAAW,CAAC,CAAC;AAExE,MAAM,WAAW,eAAe;IAC9B,EAAE,EAAE,MAAM,CAAC;IACX,QAAQ,EAAE,MAAM,CAAC;IACjB,WAAW,EAAE,MAAM,CAAC;IACpB,8FAA8F;IAC9F,SAAS,EAAE,MAAM,EAAE,EAAE,GAAG,MAAM,EAAE,CAAC;IACjC,GAAG,EAAE,aAAa,CAAC;CACpB;AAED,MAAM,WAAW,UAAU;IACzB,KAAK,CAAC,GAAG,EAAE,MAAM,EAAE,OAAO,CAAC,EAAE,kBAAkB,GAAG,OAAO,CAAC,YAAY,CAAC,CAAC;CACzE;AAED,MAAM,WAAW,kBAAkB;IACjC,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,OAAO,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IACjC,QAAQ,CAAC,EAAE,QAAQ,GAAG,QAAQ,CAAC;IAC/B,MAAM,CAAC,EAAE,WAAW,CAAC;CACtB;AAED,MAAM,WAAW,YAAY;IAC3B,EAAE,EAAE,OAAO,CAAC;IACZ,MAAM,EAAE,MAAM,CAAC;IACf,UAAU,EAAE,MAAM,CAAC;IACnB,OAAO,EAAE,OAAO,CAAC;IACjB,GAAG,EAAE,MAAM,CAAC;IACZ,UAAU,EAAE,OAAO,CAAC;IACpB,IAAI,IAAI,OAAO,CAAC,MAAM,CAAC,CAAC;CACzB;AAED,MAAM,WAAW,cAAc;IAC7B,GAAG,EAAE,MAAM,CAAC;IACZ,OAAO,EAAE,MAAM,CAAC;IAChB,MAAM,EAAE,MAAM,CAAC;IACf,UAAU,EAAE,OAAO,CAAC;IACpB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,iBAAiB,CAAC,EAAE,OAAO,CAAC;CAC7B;AAED,MAAM,WAAW,aAAc,SAAQ,YAAY;IACjD,6DAA6D;IAC7D,QAAQ,CAAC,EAAE,MAAM,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,YAAY;IAC3B,GAAG,EAAE,MAAM,CAAC;IACZ,SAAS,EAAE,MAAM,CAAC;IAClB,OAAO,EAAE,WAAW,EAAE,CAAC;IACvB,OAAO,EAAE;QACP,KAAK,EAAE,MAAM,CAAC;QACd,IAAI,EAAE,MAAM,CAAC;QACb,IAAI,EAAE,MAAM,CAAC;QACb,IAAI,EAAE,MAAM,CAAC;QACb,IAAI,EAAE,MAAM,CAAC;QACb,KAAK,EAAE,MAAM,CAAC;KACf,CAAC;CACH;AAED,MAAM,WAAW,eAAe;IAC9B,GAAG,EAAE,MAAM,CAAC;IACZ,MAAM,CAAC,EAAE,MAAM,EAAE,CAAC;IAClB,OAAO,CAAC,EAAE,OAAO,CAAC,YAAY,CAAC,CAAC;CACjC"}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "afdocs",
-  "version": "0.1.3",
+  "version": "0.2.0",
   "description": "Test your documentation site against the Agent-Friendly Documentation Spec",
   "type": "module",
   "main": "dist/index.js",
@@ -55,16 +55,18 @@
   },
   "license": "MIT",
   "engines": {
-    "node": ">=20"
+    "node": ">=22"
   },
   "dependencies": {
     "chalk": "^5.4.1",
     "commander": "^13.1.0",
+    "turndown": "^7.2.2",
     "yaml": "^2.7.0"
   },
   "devDependencies": {
     "@eslint/js": "^10.0.1",
     "@types/node": "^22.13.4",
+    "@types/turndown": "^5.0.6",
     "@vitest/coverage-v8": "^4.0.18",
     "eslint": "^10.0.1",
     "husky": "^9.1.7",