@forumulate/cli 0.9.0 → 0.10.0

package/dist/doctor/extended.js

@@ -0,0 +1,150 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.DEFAULT_SAMPLE_LIMIT = exports.DEFAULT_PLATFORM_HOST = exports.EXTENDED_CHECK_IDS = void 0;
+exports.deriveRoots = deriveRoots;
+exports.discoverPrimary = discoverPrimary;
+exports.extendedChecks = extendedChecks;
+const address_checks_1 = require("./address-checks");
+const crawl_1 = require("./crawl");
+const page_checks_1 = require("./page-checks");
+const probe_1 = require("./probe");
+const robots_params_1 = require("./robots-params");
+/**
+ * The site-wide checks added for FMU-D14, run after the single-page ones.
+ *
+ * Every one of them is a defect an integrator found on a live site after
+ * readers did, and every one is judged from the outside, by HTTP alone, so the
+ * report is the same whether it is pointed at our platform, a staging mount or
+ * any other docs host. The only thing a check knows about "the platform" is
+ * its hostname, which is an input (`--platform-host`, else the host of
+ * FORUMULATE_URL, else forumulate.com) — never a constant in the logic.
+ */
+/** Every check id this file can produce, in report order. `--only` validates against it. */
+exports.EXTENDED_CHECK_IDS = [
+    'canonical-both-addresses',
+    'sitemap-crawl',
+    'cache-consistency',
+    'root-redirect',
+    'meta-description',
+    'llms-txt',
+    'favicon',
+    'consent-tag',
+    'white-label-leaks',
+    'robots-params',
+];
+exports.DEFAULT_PLATFORM_HOST = 'forumulate.com';
+/** URLs one run fetches from the sitemaps; enough to see a pattern, few enough for CI. */
+exports.DEFAULT_SAMPLE_LIMIT = 60;
+/**
+ * A platform copy's address shape: `/<prefix>/<org>/<space>/<rest>`, where the
+ * prefix is a short reader route (`/p/`, `/b/`, `/f/` …). Read only on the
+ * platform host, where that shape is the platform's own convention.
+ */
+const PLATFORM_COPY = /^\/([a-z]{1,3})\/([^/]+)\/([^/]+)(\/.*)?$/;
+/** The roots a URL's space is served at: the mount (first segment), or the platform copy's space root. */
+function deriveRoots(url, platformHost) {
+    const u = new URL(url);
+    const path = u.pathname.replace(/\/+$/, '');
+    if ((0, probe_1.hostOf)(url) === platformHost.toLowerCase()) {
+        const m = PLATFORM_COPY.exec(path);
+        if (m)
+            return [`${u.origin}/${m[1]}/${m[2]}/${m[3]}`];
+    }
+    const first = path.split('/').filter(Boolean)[0];
+    return [first ? `${u.origin}/${first}` : u.origin];
+}
+/**
+ * The customer's address of a platform-hosted page, asked of the platform.
+ *
+ * Only for a URL on the platform host in the platform-copy shape: the public
+ * primary-domain lookup answers where the space's readers are sent, which is
+ * what both copies must name as canonical. Any failure is "unknown", and the
+ * check then compares the copies it has with each other.
+ */
+async function discoverPrimary(get, url, platformHost, platformApi) {
+    if (!platformApi || (0, probe_1.hostOf)(url) !== platformHost.toLowerCase())
+        return undefined;
+    const u = new URL(url);
+    const m = PLATFORM_COPY.exec(u.pathname.replace(/\/+$/, ''));
+    if (!m)
+        return undefined;
+    const q = new URLSearchParams({
+        org: decodeURIComponent(m[2]),
+        space: decodeURIComponent(m[3]),
+    });
+    const p = await get(`${platformApi.replace(/\/+$/, '')}/public/domains/primary?${q}`, {
+        accept: 'application/json',
+    });
+    if (!p.ok || p.status !== 200)
+        return undefined;
+    try {
+        const primary = JSON.parse(p.text).primary;
+        if (!primary?.origin)
+            return undefined;
+        const expected = `${primary.origin.replace(/\/+$/, '')}${primary.basePath ?? ''}${m[4] ?? ''}`;
+        return { expected, urls: [url, expected] };
+    }
+    catch {
+        return undefined;
+    }
+}
+async function extendedChecks(input) {
+    const want = (id) => !input.only || input.only.has(id);
+    const { get, url, html, platformHost } = input;
+    const origin = new URL(url).origin;
+    const roots = input.roots?.length
+        ? input.roots
+        : deriveRoots(input.requestedUrl ?? url, platformHost);
+    const out = [];
+    if (want('canonical-both-addresses')) {
+        const sets = [...(input.addressSets ?? [])];
+        if (sets.length === 0) {
+            // The address the user GAVE: a platform copy that redirects (the legal
+            // page went `/p/<org>/legal/<doc>` → `/legal/<org>/<doc>`) has lost the
+            // org/space shape by the time it lands.
+            const asked = input.requestedUrl ?? url;
+            const found = await discoverPrimary(get, asked, platformHost, input.platformApi);
+            if (found)
+                sets.push(found);
+        }
+        out.push(await (0, address_checks_1.canonicalBothAddresses)(get, sets));
+    }
+    // One crawl feeds three checks; it is skipped only when none of them is asked for.
+    let pages = [];
+    if (want('sitemap-crawl') || want('cache-consistency') || want('meta-description')) {
+        const crawl = await (0, crawl_1.sitemapCrawl)(get, origin, input.robots, input.limit ?? exports.DEFAULT_SAMPLE_LIMIT);
+        pages = crawl.pages;
+        if (want('sitemap-crawl'))
+            out.push(crawl.check);
+    }
+    if (want('cache-consistency'))
+        out.push((0, crawl_1.cacheConsistency)(pages));
+    if (want('root-redirect'))
+        out.push(await (0, page_checks_1.rootRedirect)(get, roots));
+    if (want('meta-description')) {
+        // The page that was asked about is judged too, even when no sitemap lists it.
+        const self = html
+            ? [{ ok: true, url, status: 200, headers: new Headers(), text: html }]
+            : [];
+        const listed = pages.some((p) => p.url === url);
+        out.push((0, page_checks_1.metaDescription)(listed ? pages : [...self, ...pages]));
+    }
+    if (want('llms-txt'))
+        out.push(await (0, page_checks_1.llmsTxt)(get, origin, roots.filter((r) => r !== origin), platformHost));
+    if (want('favicon')) {
+        out.push(html
+            ? await (0, page_checks_1.favicon)(get, url, html, roots.filter((r) => r !== origin))
+            : (0, probe_1.check)('favicon', 'The favicon resolves to an image', 'skip', 'No HTML to read.', ''));
+    }
+    if (want('consent-tag'))
+        out.push(html
+            ? (0, address_checks_1.consentTag)(html)
+            : (0, probe_1.check)('consent-tag', 'The consent banner is told the page language', 'skip', 'No HTML to read.', ''));
+    if (want('white-label-leaks'))
+        out.push(await (0, address_checks_1.whiteLabelLeaks)(get, url, html, platformHost));
+    // SEO §1: the Disallow lines for each mount's sorted / filtered copies.
+    if (want('robots-params'))
+        out.push((0, robots_params_1.robotsParams)(roots, origin, input.robots));
+    return out;
+}
+//# sourceMappingURL=extended.js.map

package/dist/doctor/extended.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"extended.js","sourceRoot":"","sources":["../../src/doctor/extended.ts"],"names":[],"mappings":";;;AAuEA,kCASC;AAUD,0CA4BC;AAED,wCAiFC;AAxMD,qDAK0B;AAC1B,mCAAyD;AACzD,+CAAgF;AAChF,mCAA8D;AAC9D,mDAA+C;AAE/C;;;;;;;;;GASG;AAEH,4FAA4F;AAC/E,QAAA,kBAAkB,GAAG;IAChC,0BAA0B;IAC1B,eAAe;IACf,mBAAmB;IACnB,eAAe;IACf,kBAAkB;IAClB,UAAU;IACV,SAAS;IACT,aAAa;IACb,mBAAmB;IACnB,eAAe;CACP,CAAC;AAEE,QAAA,qBAAqB,GAAG,gBAAgB,CAAC;AACtD,0FAA0F;AAC7E,QAAA,oBAAoB,GAAG,EAAE,CAAC;AAwBvC;;;;GAIG;AACH,MAAM,aAAa,GAAG,2CAA2C,CAAC;AAElE,0GAA0G;AAC1G,SAAgB,WAAW,CAAC,GAAW,EAAE,YAAoB;IAC3D,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC;IACvB,MAAM,IAAI,GAAG,CAAC,CAAC,QAAQ,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC;IAC5C,IAAI,IAAA,cAAM,EAAC,GAAG,CAAC,KAAK,YAAY,CAAC,WAAW,EAAE,EAAE,CAAC;QAC/C,MAAM,CAAC,GAAG,aAAa,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACnC,IAAI,CAAC;YAAE,OAAO,CAAC,GAAG,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;IACxD,CAAC;IACD,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC;IACjD,OAAO,CAAC,KAAK,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,MAAM,IAAI,KAAK,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC;AACrD,CAAC;AAED;;;;;;;GAOG;AACI,KAAK,UAAU,eAAe,CACnC,GAAQ,EACR,GAAW,EACX,YAAoB,EACpB,WAA+B;IAE/B,IAAI,CAAC,WAAW,IAAI,IAAA,cAAM,EAAC,GAAG,CAAC,KAAK,YAAY,CAAC,WAAW,EAAE;QAAE,OAAO,SAAS,CAAC;IACjF,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC;IACvB,MAAM,CAAC,GAAG,aAAa,CAAC,IAAI,CAAC,CAAC,CAAC,QAAQ,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC,CAAC;IAC7D,IAAI,CAAC,CAAC;QAAE,OAAO,SAAS,CAAC;IACzB,MAAM,CAAC,GAAG,IAAI,eAAe,CAAC;QAC5B,GAAG,EAAE,kBAAkB,CAAC,CAAC,CAAC,CAAC,CAAE,CAAC;QAC9B,KAAK,EAAE,kBAAkB,CAAC,CAAC,CAAC,CAAC,CAAE,CAAC;KACjC,CAAC,CAAC;IACH,MAAM,CAAC,GAAG,MAAM,GAAG,CAAC,GAAG,WAAW,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,2BAA2B,CAAC,EAAE,EAAE;QACpF,MAAM,EAAE,kBAAkB;KAC3B,CAAC,CAAC;IACH,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,MAAM,KAAK,GAAG;QAAE,OAAO,SAAS,CAAC;IAChD,IAAI,CAAC;QACH,MAAM,OAAO,GACX,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,IAAI,CAClB,CAAC,OAAO,CAAC;QACV,IAAI,CAAC,OAAO,EAAE,MAAM;YAAE,OAAO,SAAS,CAAC;QACvC,MAAM,QAAQ,GAAG,GAAG,OAAO,CAAC,MAAM,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,GAAG,OAAO,CAAC,QAAQ,IAAI,EAAE,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC;QAC/F,OAAO,EAAE,QAAQ,EAAE,IAAI,EAAE,CAAC,GAAG,EAAE,QAAQ,CAAC,EAAE,CAAC;IAC7C,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,SAAS,CAAC;IACnB,CAAC;AACH,CAAC;AAEM,KAAK,UAAU,cAAc,CAAC,KAAoB;IACvD,MAAM,IAAI,GAAG,CAAC,EAAU,EAAE,EAAE,CAAC,CAAC,KAAK,CAAC,IAAI,IAAI,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;IAC/D,MAAM,EAAE,GAAG,EAAE,GAAG,EAAE,IAAI,EAAE,YAAY,EAAE,GAAG,KAAK,CAAC;IAC/C,MAAM,MAAM,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC,MAAM,CAAC;IACnC,MAAM,KAAK,GAAG,KAAK,CAAC,KAAK,EAAE,MAAM;QAC/B,CAAC,CAAC,KAAK,CAAC,KAAK;QACb,CAAC,CAAC,WAAW,CAAC,KAAK,CAAC,YAAY,IAAI,GAAG,EAAE,YAAY,CAAC,CAAC;IACzD,MAAM,GAAG,GAAY,EAAE,CAAC;IAExB,IAAI,IAAI,CAAC,0BAA0B,CAAC,EAAE,CAAC;QACrC,MAAM,IAAI,GAAG,CAAC,GAAG,CAAC,KAAK,CAAC,WAAW,IAAI,EAAE,CAAC,CAAC,CAAC;QAC5C,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YACtB,uEAAuE;YACvE,wEAAwE;YACxE,wCAAwC;YACxC,MAAM,KAAK,GAAG,KAAK,CAAC,YAAY,IAAI,GAAG,CAAC;YACxC,MAAM,KAAK,GAAG,MAAM,eAAe,CAAC,GAAG,EAAE,KAAK,EAAE,YAAY,EAAE,KAAK,CAAC,WAAW,CAAC,CAAC;YACjF,IAAI,KAAK;gBAAE,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QAC9B,CAAC;QACD,GAAG,CAAC,IAAI,CAAC,MAAM,IAAA,uCAAsB,EAAC,GAAG,EAAE,IAAI,CAAC,CAAC,CAAC;IACpD,CAAC;IAED,mFAAmF;IACnF,IAAI,KAAK,GAAY,EAAE,CAAC;IACxB,IAAI,IAAI,CAAC,eAAe,CAAC,IAAI,IAAI,CAAC,mBAAmB,CAAC,IAAI,IAAI,CAAC,kBAAkB,CAAC,EAAE,CAAC;QACnF,MAAM,KAAK,GAAG,MAAM,IAAA,oBAAY,EAC9B,GAAG,EACH,MAAM,EACN,KAAK,CAAC,MAAM,EACZ,KAAK,CAAC,KAAK,IAAI,4BAAoB,CACpC,CAAC;QACF,KAAK,GAAG,KAAK,CAAC,KAAK,CAAC;QACpB,IAAI,IAAI,CAAC,eAAe,CAAC;YAAE,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC;IACnD,CAAC;IACD,IAAI,IAAI,CAAC,mBAAmB,CAAC;QAAE,GAAG,CAAC,IAAI,CAAC,IAAA,wBAAgB,EAAC,KAAK,CAAC,CAAC,CAAC;IACjE,IAAI,IAAI,CAAC,eAAe,CAAC;QAAE,GAAG,CAAC,IAAI,CAAC,MAAM,IAAA,0BAAY,EAAC,GAAG,EAAE,KAAK,CAAC,CAAC,CAAC;IACpE,IAAI,IAAI,CAAC,kBAAkB,CAAC,EAAE,CAAC;QAC7B,8EAA8E;QAC9E,MAAM,IAAI,GAAY,IAAI;YACxB,CAAC,CAAC,CAAC,EAAE,EAAE,EAAE,IAAI,EAAE,GAAG,EAAE,MAAM,EAAE,GAAG,EAAE,OAAO,EAAE,IAAI,OAAO,EAAE,EAAE,IAAI,EAAE,IAAI,EAAE,CAAC;YACtE,CAAC,CAAC,EAAE,CAAC;QACP,MAAM,MAAM,GAAG,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,GAAG,KAAK,GAAG,CAAC,CAAC;QAChD,GAAG,CAAC,IAAI,CAAC,IAAA,6BAAe,EAAC,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,GAAG,IAAI,EAAE,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;IAClE,CAAC;IACD,IAAI,IAAI,CAAC,UAAU,CAAC;QAClB,GAAG,CAAC,IAAI,CACN,MAAM,IAAA,qBAAO,EACX,GAAG,EACH,MAAM,EACN,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,KAAK,MAAM,CAAC,EACjC,YAAY,CACb,CACF,CAAC;IACJ,IAAI,IAAI,CAAC,SAAS,CAAC,EAAE,CAAC;QACpB,GAAG,CAAC,IAAI,CACN,IAAI;YACF,CAAC,CAAC,MAAM,IAAA,qBAAO,EACX,GAAG,EACH,GAAG,EACH,IAAI,EACJ,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,KAAK,MAAM,CAAC,CAClC;YACH,CAAC,CAAC,IAAA,aAAK,EAAC,SAAS,EAAE,kCAAkC,EAAE,MAAM,EAAE,kBAAkB,EAAE,EAAE,CAAC,CACzF,CAAC;IACJ,CAAC;IACD,IAAI,IAAI,CAAC,aAAa,CAAC;QACrB,GAAG,CAAC,IAAI,CACN,IAAI;YACF,CAAC,CAAC,IAAA,2BAAU,EAAC,IAAI,CAAC;YAClB,CAAC,CAAC,IAAA,aAAK,EACH,aAAa,EACb,8CAA8C,EAC9C,MAAM,EACN,kBAAkB,EAClB,EAAE,CACH,CACN,CAAC;IACJ,IAAI,IAAI,CAAC,mBAAmB,CAAC;QAAE,GAAG,CAAC,IAAI,CAAC,MAAM,IAAA,gCAAe,EAAC,GAAG,EAAE,GAAG,EAAE,IAAI,EAAE,YAAY,CAAC,CAAC,CAAC;IAC7F,wEAAwE;IACxE,IAAI,IAAI,CAAC,eAAe,CAAC;QAAE,GAAG,CAAC,IAAI,CAAC,IAAA,4BAAY,EAAC,KAAK,EAAE,MAAM,EAAE,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC;IAC/E,OAAO,GAAG,CAAC;AACb,CAAC"}

package/dist/doctor/page-checks.d.ts

@@ -0,0 +1,42 @@
+import type { Check } from '../conformance';
+import { type Get, type Probe } from './probe';
+/**
+ * `root-redirect`: a space's root either renders or moves permanently, cheaply.
+ *
+ * `GET /docs` answered `307` with a 66 KB HTML body and a canonical naming
+ * `/docs` (FMU-B06). A 307 says "this is temporary, keep asking me", so a
+ * crawler keeps the root as the indexed URL and never settles on the page it
+ * lands on; the body is bytes every visitor downloads and throws away. A 301
+ * or 308 with an empty body, or the front page rendered in place, are both fine.
+ */
+export declare function rootRedirect(get: Get, roots: string[]): Promise<Check>;
+/**
+ * `meta-description`: every page says what it is, in words somebody wrote.
+ *
+ * Checked across the crawl sample, not just the page given, because the
+ * preset fallbacks live on the pages nobody opens while testing — a space
+ * index, a blog home, a generated reference landing (FMU-B08, FMU-B09).
+ */
+export declare function metaDescription(pages: Probe[]): Check;
+/**
+ * `favicon`: the icons the page and the host advertise are images.
+ *
+ * `GET /docs/favicon.ico` on a customer's domain answered a 404 with a 104 KB
+ * HTML page (FMU-B17). Browsers, feed readers and link unfurlers ask for these
+ * addresses on their own; a 100 KB error page per ask is load, and a missing
+ * icon is a blank tab. An asset-shaped path that is not there should be a tiny
+ * 404, not the site's whole not-found page.
+ */
+export declare function favicon(get: Get, pageUrl: string, html: string, roots: string[]): Promise<Check>;
+/** One llms.txt judged against the llmstxt.org shape. */
+export declare function llmsProblems(text: string, url: string, platformHost?: string): string[];
+/**
+ * `llms-txt`: the file an AI agent reads first exists and is shaped for one.
+ *
+ * The space's llms.txt opened with "Published documentation for API Docs.",
+ * listed pages in recent-edit order with no descriptions, and titled a clash
+ * "Authentication (ref-authentication)" (FMU-B16). An agent picks pages from
+ * this list by their descriptions; without them it fetches at random.
+ */
+export declare function llmsTxt(get: Get, origin: string, roots: string[], platformHost?: string): Promise<Check>;
+//# sourceMappingURL=page-checks.d.ts.map

package/dist/doctor/page-checks.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"page-checks.d.ts","sourceRoot":"","sources":["../../src/doctor/page-checks.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,KAAK,EAAE,MAAM,gBAAgB,CAAC;AAE5C,OAAO,EAUL,KAAK,GAAG,EACR,KAAK,KAAK,EACX,MAAM,SAAS,CAAC;AAKjB;;;;;;;;GAQG;AACH,wBAAsB,YAAY,CAAC,GAAG,EAAE,GAAG,EAAE,KAAK,EAAE,MAAM,EAAE,GAAG,OAAO,CAAC,KAAK,CAAC,CA4C5E;AAED;;;;;;GAMG;AACH,wBAAgB,eAAe,CAAC,KAAK,EAAE,KAAK,EAAE,GAAG,KAAK,CAuBrD;AAID;;;;;;;;GAQG;AACH,wBAAsB,OAAO,CAC3B,GAAG,EAAE,GAAG,EACR,OAAO,EAAE,MAAM,EACf,IAAI,EAAE,MAAM,EACZ,KAAK,EAAE,MAAM,EAAE,GACd,OAAO,CAAC,KAAK,CAAC,CAyChB;AAKD,yDAAyD;AACzD,wBAAgB,YAAY,CAAC,IAAI,EAAE,MAAM,EAAE,GAAG,EAAE,MAAM,EAAE,YAAY,CAAC,EAAE,MAAM,GAAG,MAAM,EAAE,CA+BvF;AAED;;;;;;;GAOG;AACH,wBAAsB,OAAO,CAC3B,GAAG,EAAE,GAAG,EACR,MAAM,EAAE,MAAM,EACd,KAAK,EAAE,MAAM,EAAE,EACf,YAAY,CAAC,EAAE,MAAM,GACpB,OAAO,CAAC,KAAK,CAAC,CA8BhB"}

package/dist/doctor/page-checks.js

@@ -0,0 +1,200 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.rootRedirect = rootRedirect;
+exports.metaDescription = metaDescription;
+exports.favicon = favicon;
+exports.llmsProblems = llmsProblems;
+exports.llmsTxt = llmsTxt;
+const boilerplate_1 = require("./boilerplate");
+const probe_1 = require("./probe");
+/** A redirect needs a Location and nothing else; more than this is a page nobody sees. */
+const REDIRECT_BODY_LIMIT = 1024;
+/**
+ * `root-redirect`: a space's root either renders or moves permanently, cheaply.
+ *
+ * `GET /docs` answered `307` with a 66 KB HTML body and a canonical naming
+ * `/docs` (FMU-B06). A 307 says "this is temporary, keep asking me", so a
+ * crawler keeps the root as the indexed URL and never settles on the page it
+ * lands on; the body is bytes every visitor downloads and throws away. A 301
+ * or 308 with an empty body, or the front page rendered in place, are both fine.
+ */
+async function rootRedirect(get, roots) {
+    const title = 'Space roots render, or redirect permanently with no body';
+    const matters = 'A temporary (302/307) redirect keeps the root as the indexed address and the destination as a guest; a heavy redirect body is downloaded by every visitor and shown to none.';
+    const bad = [];
+    const good = [];
+    for (const root of [...new Set(roots)]) {
+        const p = await get(root, { accept: 'text/html' });
+        if (!p.ok) {
+            bad.push(`${root} → ${p.error}`);
+            continue;
+        }
+        if (p.status === 200) {
+            good.push(`${root} renders`);
+            continue;
+        }
+        if (p.status < 300 || p.status >= 400) {
+            bad.push(`${root} → ${p.status}`);
+            continue;
+        }
+        const to = p.headers.get('location') ?? '';
+        const bytes = Buffer.byteLength(p.text);
+        const problems = [];
+        if (p.status !== 301 && p.status !== 308)
+            problems.push(`${p.status} is temporary`);
+        if (bytes > REDIRECT_BODY_LIMIT)
+            problems.push(`${bytes}-byte body`);
+        if (to) {
+            const target = new URL(to, root).toString();
+            const landed = await get(target, { accept: 'text/html' });
+            const canonical = landed.ok ? (0, probe_1.canonicalOf)(landed.text, target) : undefined;
+            if (canonical && !(0, probe_1.sameUrl)(canonical, target))
+                problems.push(`its target ${target} names canonical ${canonical}`);
+        }
+        if (problems.length)
+            bad.push(`${root} → ${p.status} ${to}: ${problems.join(', ')}`);
+        else
+            good.push(`${root} → ${p.status} ${to}`);
+    }
+    if (bad.length === 0 && good.length === 0)
+        return (0, probe_1.check)('root-redirect', title, 'skip', 'No space root to check.', '');
+    return (0, probe_1.check)('root-redirect', title, bad.length ? 'fail' : 'pass', bad.length ? (0, probe_1.some)(bad) : (0, probe_1.some)(good), bad.length ? matters : '');
+}
+/**
+ * `meta-description`: every page says what it is, in words somebody wrote.
+ *
+ * Checked across the crawl sample, not just the page given, because the
+ * preset fallbacks live on the pages nobody opens while testing — a space
+ * index, a blog home, a generated reference landing (FMU-B08, FMU-B09).
+ */
+function metaDescription(pages) {
+    const title = 'Pages carry a real meta and og:description';
+    const html = pages.filter((p) => p.ok && p.status === 200);
+    if (html.length === 0)
+        return (0, probe_1.check)('meta-description', title, 'skip', 'No page answered 200 to read.', '');
+    const bad = [];
+    for (const p of html) {
+        const meta = (0, boilerplate_1.descriptionProblem)((0, probe_1.metaOf)(p.text, 'description'));
+        const og = (0, boilerplate_1.descriptionProblem)((0, probe_1.metaOf)(p.text, 'og:description'));
+        if (meta)
+            bad.push(`${p.url}: description ${meta}`);
+        else if (og)
+            bad.push(`${p.url}: og:description ${og}`);
+    }
+    return (0, probe_1.check)('meta-description', title, bad.length ? 'fail' : 'pass', bad.length
+        ? `${bad.length} of ${html.length}: ${(0, probe_1.some)(bad)}`
+        : `${html.length} page(s) described.`, bad.length
+        ? 'The description is what a search result and every link preview show. A missing one is chosen by the crawler; a template sentence tells the reader nothing about this page.'
+        : '');
+}
+const IMAGE_TYPE = /^image\/|^application\/octet-stream/i;
+/**
+ * `favicon`: the icons the page and the host advertise are images.
+ *
+ * `GET /docs/favicon.ico` on a customer's domain answered a 404 with a 104 KB
+ * HTML page (FMU-B17). Browsers, feed readers and link unfurlers ask for these
+ * addresses on their own; a 100 KB error page per ask is load, and a missing
+ * icon is a blank tab. An asset-shaped path that is not there should be a tiny
+ * 404, not the site's whole not-found page.
+ */
+async function favicon(get, pageUrl, html, roots) {
+    const title = 'The favicon resolves to an image';
+    const origin = new URL(pageUrl).origin;
+    const declared = (0, probe_1.tags)(html, 'link')
+        .filter((t) => /\bicon\b/i.test((0, probe_1.attr)(t, 'rel') ?? ''))
+        .map((t) => (0, probe_1.attr)(t, 'href'))
+        .filter((h) => !!h)
+        .map((h) => new URL(h, pageUrl).toString());
+    const required = [...new Set([`${origin}/favicon.ico`, ...declared])];
+    // Under a mount the host's own favicon is the customer's, but the mount's
+    // asset paths are ours: they must at least not answer with a page.
+    const mounted = [...new Set(roots.map((r) => `${r.replace(/\/+$/, '')}/favicon.ico`))].filter((u) => !required.includes(u));
+    const bad = [];
+    for (const url of required) {
+        const p = await get(url, { accept: 'image/*' });
+        const type = p.ok ? (p.headers.get('content-type') ?? '') : '';
+        if (!p.ok || p.status !== 200 || !IMAGE_TYPE.test(type))
+            bad.push(`${url} → ${(0, probe_1.describeProbe)(p)}${type ? ` ${type}` : ''}${p.ok ? ` (${Buffer.byteLength(p.text)} bytes)` : ''}`);
+    }
+    for (const url of mounted) {
+        const p = await get(url, { accept: 'image/*' });
+        if (!p.ok)
+            continue;
+        const type = p.headers.get('content-type') ?? '';
+        const bytes = Buffer.byteLength(p.text);
+        if (p.status === 200 && IMAGE_TYPE.test(type))
+            continue;
+        if (/html/i.test(type) && bytes > 2048)
+            bad.push(`${url} → ${p.status} with a ${bytes}-byte HTML page`);
+    }
+    return (0, probe_1.check)('favicon', title, bad.length ? 'fail' : 'pass', bad.length ? (0, probe_1.some)(bad) : `${required.length + mounted.length} icon address(es) answered.`, bad.length
+        ? 'Browsers and link previews fetch the icon on their own; an HTML error page in its place is a blank tab and wasted bytes on every request.'
+        : '');
+}
+/** `Title (some-slug)`: the generator disambiguated two titles with a slug a reader never sees. */
+const SLUG_LEAK = /\[[^\]]*\(([a-z0-9]+(?:-[a-z0-9]+)+)\)\]/;
+/** One llms.txt judged against the llmstxt.org shape. */
+function llmsProblems(text, url, platformHost) {
+    const lines = text.split(/\r?\n/);
+    const first = lines.find((l) => l.trim());
+    const out = [];
+    if (!first || !/^#\s+\S/.test(first))
+        out.push('does not open with an H1 title');
+    const quote = lines.find((l) => /^>\s*/.test(l));
+    const summary = quote?.replace(/^>\s*/, '').trim();
+    if (summary && (0, boilerplate_1.isPresetDescription)(summary))
+        out.push(`summary is preset text "${summary}"`);
+    const links = [...text.matchAll(/^\s*[-*]\s*\[([^\]]*)\]\(([^)\s]+)\)(.*)$/gm)];
+    if (links.length === 0)
+        out.push('lists no links');
+    // Links elsewhere are fine (a GitHub repo, a status page). Links back to
+    // the PLATFORM from a file on the customer's own domain are the leak: the
+    // agent is sent to a host the customer did not publish on (FMU-B15/B16).
+    const host = (0, probe_1.hostOf)(url);
+    const platform = platformHost?.toLowerCase();
+    const foreign = platform && host !== platform
+        ? links
+            .map((m) => m[2])
+            .filter((href) => /^https?:/i.test(href) && (0, probe_1.hostOf)(href)?.replace(/^www\./, '') === platform)
+        : [];
+    if (foreign.length)
+        out.push(`links to the platform host: ${(0, probe_1.some)(foreign, 3)}`);
+    const leaks = links.filter((m) => SLUG_LEAK.test(`[${m[1]}]`)).map((m) => m[1]);
+    if (leaks.length)
+        out.push(`titles carry slugs: ${(0, probe_1.some)(leaks, 3)}`);
+    const sections = lines.filter((l) => /^##\s+\S/.test(l)).length;
+    const described = links.filter((m) => /^\s*:\s*\S/.test(m[3] ?? '')).length;
+    if (links.length > 0 && sections === 0 && described === 0)
+        out.push(`${links.length} links in one flat list with no sections and no descriptions`);
+    return out;
+}
+/**
+ * `llms-txt`: the file an AI agent reads first exists and is shaped for one.
+ *
+ * The space's llms.txt opened with "Published documentation for API Docs.",
+ * listed pages in recent-edit order with no descriptions, and titled a clash
+ * "Authentication (ref-authentication)" (FMU-B16). An agent picks pages from
+ * this list by their descriptions; without them it fetches at random.
+ */
+async function llmsTxt(get, origin, roots, platformHost) {
+    const title = 'llms.txt is present and follows llmstxt.org';
+    const candidates = [
+        ...new Set([`${origin}/llms.txt`, ...roots.map((r) => `${r.replace(/\/+$/, '')}/llms.txt`)]),
+    ];
+    const found = [];
+    const bad = [];
+    for (const url of candidates) {
+        const p = await get(url, { accept: 'text/plain,text/markdown' });
+        if (!p.ok || p.status !== 200)
+            continue;
+        found.push(url);
+        for (const problem of llmsProblems(p.text, url, platformHost))
+            bad.push(`${url} ${problem}`);
+    }
+    if (found.length === 0)
+        return (0, probe_1.check)('llms-txt', title, 'fail', `None of ${candidates.join(', ')} answered 200.`, 'Agents look for /llms.txt before anything else; without it they crawl HTML and guess.');
+    return (0, probe_1.check)('llms-txt', title, bad.length ? 'fail' : 'pass', bad.length ? (0, probe_1.some)(bad) : `${found.join(', ')}: well formed.`, bad.length
+        ? 'An agent chooses which pages to read from this list; links with no descriptions, or back to the platform’s host, send it to the wrong ones.'
+        : '');
+}
+//# sourceMappingURL=page-checks.js.map

package/dist/doctor/page-checks.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"page-checks.js","sourceRoot":"","sources":["../../src/doctor/page-checks.ts"],"names":[],"mappings":";;AA4BA,oCA4CC;AASD,0CAuBC;AAaD,0BA8CC;AAMD,oCA+BC;AAUD,0BAmCC;AApPD,+CAAwE;AACxE,mCAYiB;AAEjB,0FAA0F;AAC1F,MAAM,mBAAmB,GAAG,IAAI,CAAC;AAEjC;;;;;;;;GAQG;AACI,KAAK,UAAU,YAAY,CAAC,GAAQ,EAAE,KAAe;IAC1D,MAAM,KAAK,GAAG,0DAA0D,CAAC;IACzE,MAAM,OAAO,GACX,8KAA8K,CAAC;IACjL,MAAM,GAAG,GAAa,EAAE,CAAC;IACzB,MAAM,IAAI,GAAa,EAAE,CAAC;IAC1B,KAAK,MAAM,IAAI,IAAI,CAAC,GAAG,IAAI,GAAG,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC;QACvC,MAAM,CAAC,GAAG,MAAM,GAAG,CAAC,IAAI,EAAE,EAAE,MAAM,EAAE,WAAW,EAAE,CAAC,CAAC;QACnD,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC;YACV,GAAG,CAAC,IAAI,CAAC,GAAG,IAAI,MAAM,CAAC,CAAC,KAAK,EAAE,CAAC,CAAC;YACjC,SAAS;QACX,CAAC;QACD,IAAI,CAAC,CAAC,MAAM,KAAK,GAAG,EAAE,CAAC;YACrB,IAAI,CAAC,IAAI,CAAC,GAAG,IAAI,UAAU,CAAC,CAAC;YAC7B,SAAS;QACX,CAAC;QACD,IAAI,CAAC,CAAC,MAAM,GAAG,GAAG,IAAI,CAAC,CAAC,MAAM,IAAI,GAAG,EAAE,CAAC;YACtC,GAAG,CAAC,IAAI,CAAC,GAAG,IAAI,MAAM,CAAC,CAAC,MAAM,EAAE,CAAC,CAAC;YAClC,SAAS;QACX,CAAC;QACD,MAAM,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,GAAG,CAAC,UAAU,CAAC,IAAI,EAAE,CAAC;QAC3C,MAAM,KAAK,GAAG,MAAM,CAAC,UAAU,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;QACxC,MAAM,QAAQ,GAAa,EAAE,CAAC;QAC9B,IAAI,CAAC,CAAC,MAAM,KAAK,GAAG,IAAI,CAAC,CAAC,MAAM,KAAK,GAAG;YAAE,QAAQ,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,MAAM,eAAe,CAAC,CAAC;QACpF,IAAI,KAAK,GAAG,mBAAmB;YAAE,QAAQ,CAAC,IAAI,CAAC,GAAG,KAAK,YAAY,CAAC,CAAC;QACrE,IAAI,EAAE,EAAE,CAAC;YACP,MAAM,MAAM,GAAG,IAAI,GAAG,CAAC,EAAE,EAAE,IAAI,CAAC,CAAC,QAAQ,EAAE,CAAC;YAC5C,MAAM,MAAM,GAAG,MAAM,GAAG,CAAC,MAAM,EAAE,EAAE,MAAM,EAAE,WAAW,EAAE,CAAC,CAAC;YAC1D,MAAM,SAAS,GAAG,MAAM,CAAC,EAAE,CAAC,CAAC,CAAC,IAAA,mBAAW,EAAC,MAAM,CAAC,IAAI,EAAE,MAAM,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;YAC3E,IAAI,SAAS,IAAI,CAAC,IAAA,eAAO,EAAC,SAAS,EAAE,MAAM,CAAC;gBAC1C,QAAQ,CAAC,IAAI,CAAC,cAAc,MAAM,oBAAoB,SAAS,EAAE,CAAC,CAAC;QACvE,CAAC;QACD,IAAI,QAAQ,CAAC,MAAM;YAAE,GAAG,CAAC,IAAI,CAAC,GAAG,IAAI,MAAM,CAAC,CAAC,MAAM,IAAI,EAAE,KAAK,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;;YAChF,IAAI,CAAC,IAAI,CAAC,GAAG,IAAI,MAAM,CAAC,CAAC,MAAM,IAAI,EAAE,EAAE,CAAC,CAAC;IAChD,CAAC;IACD,IAAI,GAAG,CAAC,MAAM,KAAK,CAAC,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC;QACvC,OAAO,IAAA,aAAK,EAAC,eAAe,EAAE,KAAK,EAAE,MAAM,EAAE,yBAAyB,EAAE,EAAE,CAAC,CAAC;IAC9E,OAAO,IAAA,aAAK,EACV,eAAe,EACf,KAAK,EACL,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,EAC5B,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,IAAA,YAAI,EAAC,GAAG,CAAC,CAAC,CAAC,CAAC,IAAA,YAAI,EAAC,IAAI,CAAC,EACnC,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,CAC1B,CAAC;AACJ,CAAC;AAED;;;;;;GAMG;AACH,SAAgB,eAAe,CAAC,KAAc;IAC5C,MAAM,KAAK,GAAG,4CAA4C,CAAC;IAC3D,MAAM,IAAI,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAqC,EAAE,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,MAAM,KAAK,GAAG,CAAC,CAAC;IAC9F,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC;QACnB,OAAO,IAAA,aAAK,EAAC,kBAAkB,EAAE,KAAK,EAAE,MAAM,EAAE,+BAA+B,EAAE,EAAE,CAAC,CAAC;IACvF,MAAM,GAAG,GAAa,EAAE,CAAC;IACzB,KAAK,MAAM,CAAC,IAAI,IAAI,EAAE,CAAC;QACrB,MAAM,IAAI,GAAG,IAAA,gCAAkB,EAAC,IAAA,cAAM,EAAC,CAAC,CAAC,IAAI,EAAE,aAAa,CAAC,CAAC,CAAC;QAC/D,MAAM,EAAE,GAAG,IAAA,gCAAkB,EAAC,IAAA,cAAM,EAAC,CAAC,CAAC,IAAI,EAAE,gBAAgB,CAAC,CAAC,CAAC;QAChE,IAAI,IAAI;YAAE,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,GAAG,iBAAiB,IAAI,EAAE,CAAC,CAAC;aAC/C,IAAI,EAAE;YAAE,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,GAAG,oBAAoB,EAAE,EAAE,CAAC,CAAC;IAC1D,CAAC;IACD,OAAO,IAAA,aAAK,EACV,kBAAkB,EAClB,KAAK,EACL,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,EAC5B,GAAG,CAAC,MAAM;QACR,CAAC,CAAC,GAAG,GAAG,CAAC,MAAM,OAAO,IAAI,CAAC,MAAM,KAAK,IAAA,YAAI,EAAC,GAAG,CAAC,EAAE;QACjD,CAAC,CAAC,GAAG,IAAI,CAAC,MAAM,qBAAqB,EACvC,GAAG,CAAC,MAAM;QACR,CAAC,CAAC,4KAA4K;QAC9K,CAAC,CAAC,EAAE,CACP,CAAC;AACJ,CAAC;AAED,MAAM,UAAU,GAAG,sCAAsC,CAAC;AAE1D;;;;;;;;GAQG;AACI,KAAK,UAAU,OAAO,CAC3B,GAAQ,EACR,OAAe,EACf,IAAY,EACZ,KAAe;IAEf,MAAM,KAAK,GAAG,kCAAkC,CAAC;IACjD,MAAM,MAAM,GAAG,IAAI,GAAG,CAAC,OAAO,CAAC,CAAC,MAAM,CAAC;IACvC,MAAM,QAAQ,GAAG,IAAA,YAAI,EAAC,IAAI,EAAE,MAAM,CAAC;SAChC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,WAAW,CAAC,IAAI,CAAC,IAAA,YAAI,EAAC,CAAC,EAAE,KAAK,CAAC,IAAI,EAAE,CAAC,CAAC;SACrD,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,IAAA,YAAI,EAAC,CAAC,EAAE,MAAM,CAAC,CAAC;SAC3B,MAAM,CAAC,CAAC,CAAC,EAAe,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;SAC/B,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,IAAI,GAAG,CAAC,CAAC,EAAE,OAAO,CAAC,CAAC,QAAQ,EAAE,CAAC,CAAC;IAC9C,MAAM,QAAQ,GAAG,CAAC,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,MAAM,cAAc,EAAE,GAAG,QAAQ,CAAC,CAAC,CAAC,CAAC;IACtE,0EAA0E;IAC1E,mEAAmE;IACnE,MAAM,OAAO,GAAG,CAAC,GAAG,IAAI,GAAG,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,CAAC,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,cAAc,CAAC,CAAC,CAAC,CAAC,MAAM,CAC3F,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,QAAQ,CAAC,QAAQ,CAAC,CAAC,CAAC,CAC7B,CAAC;IACF,MAAM,GAAG,GAAa,EAAE,CAAC;IACzB,KAAK,MAAM,GAAG,IAAI,QAAQ,EAAE,CAAC;QAC3B,MAAM,CAAC,GAAG,MAAM,GAAG,CAAC,GAAG,EAAE,EAAE,MAAM,EAAE,SAAS,EAAE,CAAC,CAAC;QAChD,MAAM,IAAI,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,GAAG,CAAC,cAAc,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;QAC/D,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,MAAM,KAAK,GAAG,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC;YACrD,GAAG,CAAC,IAAI,CACN,GAAG,GAAG,MAAM,IAAA,qBAAa,EAAC,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC,CAAC,IAAI,IAAI,EAAE,CAAC,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,KAAK,MAAM,CAAC,UAAU,CAAC,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC,CAAC,EAAE,EAAE,CAC9G,CAAC;IACN,CAAC;IACD,KAAK,MAAM,GAAG,IAAI,OAAO,EAAE,CAAC;QAC1B,MAAM,CAAC,GAAG,MAAM,GAAG,CAAC,GAAG,EAAE,EAAE,MAAM,EAAE,SAAS,EAAE,CAAC,CAAC;QAChD,IAAI,CAAC,CAAC,CAAC,EAAE;YAAE,SAAS;QACpB,MAAM,IAAI,GAAG,CAAC,CAAC,OAAO,CAAC,GAAG,CAAC,cAAc,CAAC,IAAI,EAAE,CAAC;QACjD,MAAM,KAAK,GAAG,MAAM,CAAC,UAAU,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;QACxC,IAAI,CAAC,CAAC,MAAM,KAAK,GAAG,IAAI,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC;YAAE,SAAS;QACxD,IAAI,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,KAAK,GAAG,IAAI;YACpC,GAAG,CAAC,IAAI,CAAC,GAAG,GAAG,MAAM,CAAC,CAAC,MAAM,WAAW,KAAK,iBAAiB,CAAC,CAAC;IACpE,CAAC;IACD,OAAO,IAAA,aAAK,EACV,SAAS,EACT,KAAK,EACL,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,EAC5B,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,IAAA,YAAI,EAAC,GAAG,CAAC,CAAC,CAAC,CAAC,GAAG,QAAQ,CAAC,MAAM,GAAG,OAAO,CAAC,MAAM,6BAA6B,EACzF,GAAG,CAAC,MAAM;QACR,CAAC,CAAC,2IAA2I;QAC7I,CAAC,CAAC,EAAE,CACP,CAAC;AACJ,CAAC;AAED,mGAAmG;AACnG,MAAM,SAAS,GAAG,0CAA0C,CAAC;AAE7D,yDAAyD;AACzD,SAAgB,YAAY,CAAC,IAAY,EAAE,GAAW,EAAE,YAAqB;IAC3E,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;IAClC,MAAM,KAAK,GAAG,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC;IAC1C,MAAM,GAAG,GAAa,EAAE,CAAC;IACzB,IAAI,CAAC,KAAK,IAAI,CAAC,SAAS,CAAC,IAAI,CAAC,KAAK,CAAC;QAAE,GAAG,CAAC,IAAI,CAAC,gCAAgC,CAAC,CAAC;IACjF,MAAM,KAAK,GAAG,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;IACjD,MAAM,OAAO,GAAG,KAAK,EAAE,OAAO,CAAC,OAAO,EAAE,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC;IACnD,IAAI,OAAO,IAAI,IAAA,iCAAmB,EAAC,OAAO,CAAC;QAAE,GAAG,CAAC,IAAI,CAAC,2BAA2B,OAAO,GAAG,CAAC,CAAC;IAC7F,MAAM,KAAK,GAAG,CAAC,GAAG,IAAI,CAAC,QAAQ,CAAC,6CAA6C,CAAC,CAAC,CAAC;IAChF,IAAI,KAAK,CAAC,MAAM,KAAK,CAAC;QAAE,GAAG,CAAC,IAAI,CAAC,gBAAgB,CAAC,CAAC;IACnD,yEAAyE;IACzE,0EAA0E;IAC1E,yEAAyE;IACzE,MAAM,IAAI,GAAG,IAAA,cAAM,EAAC,GAAG,CAAC,CAAC;IACzB,MAAM,QAAQ,GAAG,YAAY,EAAE,WAAW,EAAE,CAAC;IAC7C,MAAM,OAAO,GACX,QAAQ,IAAI,IAAI,KAAK,QAAQ;QAC3B,CAAC,CAAC,KAAK;aACF,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAE,CAAC;aACjB,MAAM,CACL,CAAC,IAAI,EAAE,EAAE,CAAC,WAAW,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,IAAA,cAAM,EAAC,IAAI,CAAC,EAAE,OAAO,CAAC,QAAQ,EAAE,EAAE,CAAC,KAAK,QAAQ,CACrF;QACL,CAAC,CAAC,EAAE,CAAC;IACT,IAAI,OAAO,CAAC,MAAM;QAAE,GAAG,CAAC,IAAI,CAAC,+BAA+B,IAAA,YAAI,EAAC,OAAO,EAAE,CAAC,CAAC,EAAE,CAAC,CAAC;IAChF,MAAM,KAAK,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,SAAS,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAE,CAAC,CAAC;IACjF,IAAI,KAAK,CAAC,MAAM;QAAE,GAAG,CAAC,IAAI,CAAC,uBAAuB,IAAA,YAAI,EAAC,KAAK,EAAE,CAAC,CAAC,EAAE,CAAC,CAAC;IACpE,MAAM,QAAQ,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,UAAU,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC;IAChE,MAAM,SAAS,GAAG,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,YAAY,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC;IAC5E,IAAI,KAAK,CAAC,MAAM,GAAG,CAAC,IAAI,QAAQ,KAAK,CAAC,IAAI,SAAS,KAAK,CAAC;QACvD,GAAG,CAAC,IAAI,CAAC,GAAG,KAAK,CAAC,MAAM,8DAA8D,CAAC,CAAC;IAC1F,OAAO,GAAG,CAAC;AACb,CAAC;AAED;;;;;;;GAOG;AACI,KAAK,UAAU,OAAO,CAC3B,GAAQ,EACR,MAAc,EACd,KAAe,EACf,YAAqB;IAErB,MAAM,KAAK,GAAG,6CAA6C,CAAC;IAC5D,MAAM,UAAU,GAAG;QACjB,GAAG,IAAI,GAAG,CAAC,CAAC,GAAG,MAAM,WAAW,EAAE,GAAG,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,CAAC,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,WAAW,CAAC,CAAC,CAAC;KAC7F,CAAC;IACF,MAAM,KAAK,GAAa,EAAE,CAAC;IAC3B,MAAM,GAAG,GAAa,EAAE,CAAC;IACzB,KAAK,MAAM,GAAG,IAAI,UAAU,EAAE,CAAC;QAC7B,MAAM,CAAC,GAAG,MAAM,GAAG,CAAC,GAAG,EAAE,EAAE,MAAM,EAAE,0BAA0B,EAAE,CAAC,CAAC;QACjE,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,MAAM,KAAK,GAAG;YAAE,SAAS;QACxC,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QAChB,KAAK,MAAM,OAAO,IAAI,YAAY,CAAC,CAAC,CAAC,IAAI,EAAE,GAAG,EAAE,YAAY,CAAC;YAAE,GAAG,CAAC,IAAI,CAAC,GAAG,GAAG,IAAI,OAAO,EAAE,CAAC,CAAC;IAC/F,CAAC;IACD,IAAI,KAAK,CAAC,MAAM,KAAK,CAAC;QACpB,OAAO,IAAA,aAAK,EACV,UAAU,EACV,KAAK,EACL,MAAM,EACN,WAAW,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC,gBAAgB,EAChD,uFAAuF,CACxF,CAAC;IACJ,OAAO,IAAA,aAAK,EACV,UAAU,EACV,KAAK,EACL,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,MAAM,EAC5B,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,IAAA,YAAI,EAAC,GAAG,CAAC,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,gBAAgB,EAC5D,GAAG,CAAC,MAAM;QACR,CAAC,CAAC,6IAA6I;QAC/I,CAAC,CAAC,EAAE,CACP,CAAC;AACJ,CAAC"}

package/dist/doctor/probe.d.ts

@@ -0,0 +1,63 @@
+import type { Check } from '../conformance';
+/**
+ * The HTTP and HTML plumbing every doctor check shares.
+ *
+ * One `get`, so every request the doctor makes has the same three properties:
+ * it never follows a redirect on its own (a redirect is usually the thing being
+ * checked, and `fetch`'s default hides both the status and the body), it gives
+ * up after a fixed time (one hung address must not hang a CI job), and it never
+ * throws (a refused connection is a finding, not a crash).
+ */
+export type Probe = {
+    ok: true;
+    url: string;
+    status: number;
+    headers: Headers;
+    text: string;
+} | {
+    ok: false;
+    url: string;
+    error: string;
+};
+export interface ProbeOptions {
+    fetchImpl: typeof fetch;
+    /** Per-request budget. A crawl of a hundred pages must not wait on one forever. */
+    timeoutMs?: number;
+}
+export declare const DEFAULT_TIMEOUT_MS = 15000;
+export declare function makeGet(opts: ProbeOptions): (url: string, headers?: Record<string, string>) => Promise<Probe>;
+export type Get = ReturnType<typeof makeGet>;
+/**
+ * Run `fn` over `items`, at most `limit` at a time, in input order.
+ *
+ * Bounded on purpose: the doctor runs from a laptop or a CI runner against a
+ * customer's production site, and a hundred simultaneous requests is how a
+ * health check becomes the incident it was meant to catch.
+ */
+export declare function pool<T, R>(items: T[], limit: number, fn: (t: T) => Promise<R>): Promise<R[]>;
+/** Follow redirects by hand (bounded), keeping every hop — the hops are what a check judges. */
+export declare function follow(get: Get, url: string, headers?: Record<string, string>, maxHops?: number): Promise<{
+    hops: Probe[];
+    final: Probe;
+}>;
+/** One attribute's value from one tag's source, quoted or not. */
+declare function attrOf(tag: string, name: string): string | undefined;
+export declare function tags(html: string, name: string): string[];
+export declare const attr: typeof attrOf;
+/** `<link rel="canonical">`, absolute against `base`, or undefined. */
+export declare function canonicalOf(html: string, base: string): string | undefined;
+/** `<meta name|property=X content=…>`. */
+export declare function metaOf(html: string, key: string): string | undefined;
+/** The document's declared language (`<html lang>`), lower-cased, or undefined. */
+export declare function htmlLang(html: string): string | undefined;
+export declare function decodeEntities(s: string): string;
+export declare function hostOf(href: string): string | undefined;
+/** Compare two URLs the way a search engine would: no fragment, no trailing slash. */
+export declare function sameUrl(a: string | undefined, b: string | undefined): boolean;
+export declare function describeProbe(p: Probe): string;
+/** A check under construction: the shape `runDoctor` reports. */
+export declare function check(id: string, title: string, state: Check['state'], detail: string, matters?: string): Check;
+/** At most `n` items of a list in a detail line, and how many were left out. */
+export declare function some(list: string[], n?: number): string;
+export {};
+//# sourceMappingURL=probe.d.ts.map

package/dist/doctor/probe.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"probe.d.ts","sourceRoot":"","sources":["../../src/doctor/probe.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,KAAK,EAAE,MAAM,gBAAgB,CAAC;AAE5C;;;;;;;;GAQG;AACH,MAAM,MAAM,KAAK,GACb;IAAE,EAAE,EAAE,IAAI,CAAC;IAAC,GAAG,EAAE,MAAM,CAAC;IAAC,MAAM,EAAE,MAAM,CAAC;IAAC,OAAO,EAAE,OAAO,CAAC;IAAC,IAAI,EAAE,MAAM,CAAA;CAAE,GACzE;IAAE,EAAE,EAAE,KAAK,CAAC;IAAC,GAAG,EAAE,MAAM,CAAC;IAAC,KAAK,EAAE,MAAM,CAAA;CAAE,CAAC;AAE9C,MAAM,WAAW,YAAY;IAC3B,SAAS,EAAE,OAAO,KAAK,CAAC;IACxB,mFAAmF;IACnF,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAED,eAAO,MAAM,kBAAkB,QAAS,CAAC;AAEzC,wBAAgB,OAAO,CAAC,IAAI,EAAE,YAAY,IAE1B,KAAK,MAAM,EAAE,UAAS,MAAM,CAAC,MAAM,EAAE,MAAM,CAAM,KAAG,OAAO,CAAC,KAAK,CAAC,CA0BjF;AAED,MAAM,MAAM,GAAG,GAAG,UAAU,CAAC,OAAO,OAAO,CAAC,CAAC;AAE7C;;;;;;GAMG;AACH,wBAAsB,IAAI,CAAC,CAAC,EAAE,CAAC,EAC7B,KAAK,EAAE,CAAC,EAAE,EACV,KAAK,EAAE,MAAM,EACb,EAAE,EAAE,CAAC,CAAC,EAAE,CAAC,KAAK,OAAO,CAAC,CAAC,CAAC,GACvB,OAAO,CAAC,CAAC,EAAE,CAAC,CAWd;AAED,gGAAgG;AAChG,wBAAsB,MAAM,CAC1B,GAAG,EAAE,GAAG,EACR,GAAG,EAAE,MAAM,EACX,OAAO,GAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAM,EACpC,OAAO,SAAI,GACV,OAAO,CAAC;IAAE,IAAI,EAAE,KAAK,EAAE,CAAC;IAAC,KAAK,EAAE,KAAK,CAAA;CAAE,CAAC,CAY1C;AAED,kEAAkE;AAClE,iBAAS,MAAM,CAAC,GAAG,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS,CAG7D;AAED,wBAAgB,IAAI,CAAC,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,GAAG,MAAM,EAAE,CAEzD;AAED,eAAO,MAAM,IAAI,eAAS,CAAC;AAE3B,uEAAuE;AACvE,wBAAgB,WAAW,CAAC,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS,CAa1E;AAED,0CAA0C;AAC1C,wBAAgB,MAAM,CAAC,IAAI,EAAE,MAAM,EAAE,GAAG,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS,CAMpE;AAED,mFAAmF;AACnF,wBAAgB,QAAQ,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS,CAGzD;AAED,wBAAgB,cAAc,CAAC,CAAC,EAAE,MAAM,GAAG,MAAM,CAOhD;AAED,wBAAgB,MAAM,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,GAAG,SAAS,CAMvD;AAED,sFAAsF;AACtF,wBAAgB,OAAO,CAAC,CAAC,EAAE,MAAM,GAAG,SAAS,EAAE,CAAC,EAAE,MAAM,GAAG,SAAS,GAAG,OAAO,CAY7E;AAED,wBAAgB,aAAa,CAAC,CAAC,EAAE,KAAK,GAAG,MAAM,CAE9C;AAED,iEAAiE;AACjE,wBAAgB,KAAK,CACnB,EAAE,EAAE,MAAM,EACV,KAAK,EAAE,MAAM,EACb,KAAK,EAAE,KAAK,CAAC,OAAO,CAAC,EACrB,MAAM,EAAE,MAAM,EACd,OAAO,SAAK,GACX,KAAK,CAEP;AAED,gFAAgF;AAChF,wBAAgB,IAAI,CAAC,IAAI,EAAE,MAAM,EAAE,EAAE,CAAC,SAAI,GAAG,MAAM,CAGlD"}