@canonry/canonry 4.162.0 → 4.163.0

package/dist/{chunk-FZAAGXTK.js → chunk-MRHYSEZB.js}

@@ -10,7 +10,7 @@ import {
   loadConfig,
   loadConfigRaw,
   saveConfigPatch
-} from "./chunk-2TLRE2L3.js";
+} from "./chunk-INXI6PO7.js";
 import {
   CC_CACHE_DIR,
   DUCKDB_SPEC,
@@ -139,7 +139,7 @@ import {
   siteCrawlSnapshots,
   toAlertView,
   usageCounters
-} from "./chunk-C6ZZH3XI.js";
+} from "./chunk-G5UZTODF.js";
 import {
   AGENT_MEMORY_VALUE_MAX_BYTES,
   AGENT_PROVIDER_IDS,
@@ -172,7 +172,7 @@ import {
   SITE_CRAWL_GRAPH_MAX_NODES,
   SKILL_MANIFEST_FILENAME,
   SchedulableRunKinds,
-  SiteHealthTemplateDetections,
+  SiteHealthLinkClassificationSources,
   SkillsClients,
   TEMPLATE_LINK_RATIO_THRESHOLD,
   TrafficEventConfidences,
@@ -193,11 +193,13 @@ import {
   bucketOnboardingCount,
   buildRunErrorFromMessages,
   classifySkillFile,
+  classifyTemplateLinkEdge,
   cloudflareQueueNameSchema,
   coerceSkillManifest,
   computeBacklinkSummaryMetrics,
   contentActionLabel,
   contentBriefDtoSchema,
+  createTemplateLinkDetectionTally,
   createTemplateLinkPairIndex,
   deriveCitedUrlCandidates,
   deriveIndexCoverage,
@@ -220,7 +222,6 @@ import {
   isGhostTelemetryEvent,
   isRetryableHttpError,
   isSelfLink,
-  isTemplateLinkRatio,
   isVertexGroundingRedirect,
   mapWithConcurrency,
   normalizeIdTokens,
@@ -229,6 +230,7 @@ import {
   normalizeSiteAuditRunRequest,
   normalizeUrlPath,
   notFound,
+  observeTemplateLinkDetection,
   observeTemplateLinkEdges,
   parseMeasurementRunManifestV1,
   parseOriginList,
@@ -244,13 +246,14 @@ import {
   splitList,
   startOfDayHourInTimeZone,
   templateLinkDetection,
-  templateLinkRatio,
+  templateLinkPlacementAvailable,
+  templateLinkUbiquityAvailable,
   textContainsBrandAlias,
   textContainsDomain,
   validationError,
   winnabilityClassLabel,
   withRetry
-} from "./chunk-NG7OR6GO.js";
+} from "./chunk-ESH3LPYQ.js";
 
 // src/telemetry.ts
 import crypto from "crypto";
@@ -7304,7 +7307,7 @@ var COMPONENT_PACKING_SPACING_FACTOR = 3;
 function siteCrawlGraphLayoutSettingsFingerprint(settings) {
   return crypto16.createHash("sha256").update(JSON.stringify(settings)).digest("hex").slice(0, 8);
 }
-var SITE_CRAWL_GRAPH_LAYOUT_ALGORITHM = "site-health-fa2-v4";
+var SITE_CRAWL_GRAPH_LAYOUT_ALGORITHM = "site-health-fa2-v5";
 var SITE_CRAWL_GRAPH_LAYOUT_VERSION = `${SITE_CRAWL_GRAPH_LAYOUT_ALGORITHM}-${siteCrawlGraphLayoutSettingsFingerprint({
   forceAtlas: FORCE_ATLAS_SETTINGS,
   noverlap: NOVERLAP_SETTINGS,
@@ -7860,6 +7863,13 @@ function deleteSiteCrawlGraphLayout(db, scope) {
 import { and as and11, asc as asc2, eq as eq14, gt, inArray as inArray4 } from "drizzle-orm";
 var TEMPLATE_LINK_READ_BATCH = 2e3;
 var TEMPLATE_LINK_WRITE_BATCH = 500;
+function placementOccurrencesOf(row) {
+  const { placementNavigationOccurrences: navigation } = row;
+  const { placementContentOccurrences: content } = row;
+  const { placementUnknownOccurrences: unknown } = row;
+  if (navigation == null || content == null || unknown == null) return null;
+  return { navigation, content, unknown };
+}
 function* streamAttemptEdges(db, scope) {
   let after = "";
   for (; ; ) {
@@ -7867,7 +7877,11 @@ function* streamAttemptEdges(db, scope) {
       edgeKey: siteCrawlEdges.edgeKey,
       sourceNodeKey: siteCrawlEdges.sourceNodeKey,
       targetNodeKey: siteCrawlEdges.targetNodeKey,
-      anchors: siteCrawlEdges.anchors
+      anchors: siteCrawlEdges.anchors,
+      relation: siteCrawlEdges.relation,
+      placementNavigationOccurrences: siteCrawlEdges.placementNavigationOccurrences,
+      placementContentOccurrences: siteCrawlEdges.placementContentOccurrences,
+      placementUnknownOccurrences: siteCrawlEdges.placementUnknownOccurrences
     }).from(siteCrawlEdges).where(and11(
       eq14(siteCrawlEdges.projectId, scope.projectId),
       eq14(siteCrawlEdges.runId, scope.runId),
@@ -7875,7 +7889,14 @@ function* streamAttemptEdges(db, scope) {
       gt(siteCrawlEdges.edgeKey, after)
     )).orderBy(asc2(siteCrawlEdges.edgeKey)).limit(TEMPLATE_LINK_READ_BATCH).all();
     if (rows.length === 0) return;
-    yield rows;
+    yield rows.map((row) => ({
+      edgeKey: row.edgeKey,
+      sourceNodeKey: row.sourceNodeKey,
+      targetNodeKey: row.targetNodeKey,
+      anchors: row.anchors,
+      relation: row.relation,
+      placementOccurrences: placementOccurrencesOf(row)
+    }));
     after = rows[rows.length - 1].edgeKey;
     if (rows.length < TEMPLATE_LINK_READ_BATCH) return;
   }
@@ -7885,39 +7906,56 @@ function* chunked(values, size = TEMPLATE_LINK_WRITE_BATCH) {
     yield values.slice(offset, offset + size);
   }
 }
-function classifySiteCrawlTemplateLinks(db, scope, pagesFetched) {
+function writeGroupKey(isTemplate, templateRatio) {
+  return `${isTemplate ? "1" : "0"}${templateRatio ?? ""}`;
+}
+function classifySiteCrawlTemplateLinks(db, scope, pagesFetched, placementRulesetVersion) {
   const scopeFilter = and11(
     eq14(siteCrawlEdges.projectId, scope.projectId),
     eq14(siteCrawlEdges.runId, scope.runId),
     eq14(siteCrawlEdges.attemptId, scope.attemptId)
   );
-  const detection = templateLinkDetection(pagesFetched);
+  const placementAvailable = templateLinkPlacementAvailable(placementRulesetVersion);
+  const ubiquityAvailable = templateLinkUbiquityAvailable(pagesFetched);
   db.update(siteCrawlEdges).set({ isTemplate: false, templateRatio: null }).where(scopeFilter).run();
-  if (detection !== SiteHealthTemplateDetections.applied) {
-    return { detection, templateEdgeCount: 0 };
-  }
   const index = createTemplateLinkPairIndex();
-  for (const batch of streamAttemptEdges(db, scope)) observeTemplateLinkEdges(index, batch);
-  const edgeKeysByRatio = /* @__PURE__ */ new Map();
+  if (ubiquityAvailable) {
+    for (const batch of streamAttemptEdges(db, scope)) observeTemplateLinkEdges(index, batch);
+  }
+  const edgeKeysByValue = /* @__PURE__ */ new Map();
   let templateEdgeCount = 0;
+  let placementEdgeCount = 0;
+  const tally = createTemplateLinkDetectionTally();
   for (const batch of streamAttemptEdges(db, scope)) {
     for (const edge of batch) {
-      const ratio = templateLinkRatio(index, pagesFetched, edge);
-      if (ratio == null) continue;
-      const group = edgeKeysByRatio.get(ratio);
-      if (group) group.push(edge.edgeKey);
-      else edgeKeysByRatio.set(ratio, [edge.edgeKey]);
-      if (isTemplateLinkRatio(ratio)) templateEdgeCount += 1;
+      const decision = classifyTemplateLinkEdge(edge, { index, pagesFetched, placementAvailable, ubiquityAvailable });
+      observeTemplateLinkDetection(tally, edge, decision);
+      if (decision.source === SiteHealthLinkClassificationSources.placement) placementEdgeCount += 1;
+      if (decision.isTemplate) templateEdgeCount += 1;
+      if (!decision.isTemplate && decision.templateRatio == null) continue;
+      const key = writeGroupKey(decision.isTemplate, decision.templateRatio);
+      const group = edgeKeysByValue.get(key);
+      if (group) group.edgeKeys.push(decision.edgeKey);
+      else {
+        edgeKeysByValue.set(key, {
+          isTemplate: decision.isTemplate,
+          templateRatio: decision.templateRatio,
+          edgeKeys: [decision.edgeKey]
+        });
+      }
     }
   }
   const now = (/* @__PURE__ */ new Date()).toISOString();
-  for (const [ratio, edgeKeys] of edgeKeysByRatio) {
-    const isTemplate = isTemplateLinkRatio(ratio);
-    for (const chunk of chunked(edgeKeys)) {
-      db.update(siteCrawlEdges).set({ isTemplate, templateRatio: ratio, updatedAt: now }).where(and11(scopeFilter, inArray4(siteCrawlEdges.edgeKey, chunk))).run();
+  for (const group of edgeKeysByValue.values()) {
+    for (const chunk of chunked(group.edgeKeys)) {
+      db.update(siteCrawlEdges).set({ isTemplate: group.isTemplate, templateRatio: group.templateRatio, updatedAt: now }).where(and11(scopeFilter, inArray4(siteCrawlEdges.edgeKey, chunk))).run();
     }
   }
-  return { detection, templateEdgeCount };
+  return {
+    detection: templateLinkDetection({ placementAvailable, ubiquityAvailable, tally }),
+    templateEdgeCount,
+    placementEdgeCount
+  };
 }
 
 // src/execute-site-audit.ts
@@ -8230,6 +8268,11 @@ async function executeSiteAudit(db, runId, projectId, opts = {}) {
     for (const page of pages) persistPage(tx, page, now);
     bindReferences(tx, changedUrls, now);
   };
+  const edgePlacementColumns = (edge) => ({
+    placementNavigationOccurrences: edge.placementOccurrences?.navigation ?? null,
+    placementContentOccurrences: edge.placementOccurrences?.content ?? null,
+    placementUnknownOccurrences: edge.placementOccurrences?.unknown ?? null
+  });
   const persistEdge = (tx, edge, now) => {
     if (isSelfLink(edge.from, edge.to)) return;
     observedEdges.set(edge.key, edge);
@@ -8250,6 +8293,7 @@ async function executeSiteAudit(db, runId, projectId, opts = {}) {
       followableOccurrences: edge.followableOccurrences,
       nofollowOccurrences: edge.nofollowOccurrences,
       anchors: edge.anchorSummaries.map((anchor) => anchor.text),
+      ...edgePlacementColumns(edge),
       createdAt: now,
       updatedAt: now
     }).onConflictDoUpdate({
@@ -8266,6 +8310,7 @@ async function executeSiteAudit(db, runId, projectId, opts = {}) {
         followableOccurrences: edge.followableOccurrences,
         nofollowOccurrences: edge.nofollowOccurrences,
         anchors: edge.anchorSummaries.map((anchor) => anchor.text),
+        ...edgePlacementColumns(edge),
         updatedAt: now
       }
     }).run();
@@ -8373,16 +8418,20 @@ async function executeSiteAudit(db, runId, projectId, opts = {}) {
     if (crawlSummary.complete && !hasAuditedPages) {
       throw emptyCompleteCrawlError(crawlSummary.rootUrl, crawlSummary.finalRootUrl, crawlSummary.pagesObserved);
     }
+    const placementRulesetVersion = crawlSummary.linkPlacementRulesetVersion ?? null;
     const templateLinks = classifySiteCrawlTemplateLinks(
       db,
       { projectId, runId, attemptId },
-      crawlSummary.pagesFetched
+      crawlSummary.pagesFetched,
+      placementRulesetVersion
     );
     log12.info("template-links", {
       runId,
       projectId,
       detection: templateLinks.detection,
-      templateEdges: templateLinks.templateEdgeCount
+      templateEdges: templateLinks.templateEdgeCount,
+      placementEdges: templateLinks.placementEdgeCount,
+      placementRulesetVersion
     });
     const graphLayout = await prepareSiteCrawlGraphLayout(db, {
       projectId,
@@ -8497,6 +8546,7 @@ async function executeSiteAudit(db, runId, projectId, opts = {}) {
         deadLinksChecked,
         deadLinksFound,
         templateDetection: templateLinks.detection,
+        linkPlacementRulesetVersion: placementRulesetVersion,
         createdAt: finishedAt,
         updatedAt: finishedAt
       }).run();
@@ -9034,7 +9084,7 @@ function readStoredGroundingSources(rawResponse) {
   return result;
 }
 async function backfillInsightsCommand(project, opts) {
-  const { IntelligenceService: IntelligenceService2 } = await import("./intelligence-service-ANRFEQWX.js");
+  const { IntelligenceService: IntelligenceService2 } = await import("./intelligence-service-AFAFGLCM.js");
   const config = loadConfig();
   const db = createClient(config.database);
   migrate(db);

package/dist/cli.js

@@ -31,7 +31,7 @@ import {
   showFirstRunNotice,
   trackCliCommandFinished,
   trackEvent
-} from "./chunk-FZAAGXTK.js";
+} from "./chunk-MRHYSEZB.js";
 import {
   CliError,
   EXIT_SYSTEM_ERROR,
@@ -51,7 +51,7 @@ import {
   saveConfigPatch,
   systemError,
   usageError
-} from "./chunk-2TLRE2L3.js";
+} from "./chunk-INXI6PO7.js";
 import {
   CLOUDFLARE_WORKER_BINDINGS,
   CLOUDFLARE_WORKER_GENERATED_MARKER,
@@ -65,7 +65,7 @@ import {
   projects,
   queries,
   renderReportHtml
-} from "./chunk-C6ZZH3XI.js";
+} from "./chunk-G5UZTODF.js";
 import {
   AdsDeliverySnapshotStatuses,
   AdsHistoricalCampaignRollupStatuses,
@@ -129,7 +129,7 @@ import {
   providerQuotaPolicySchema,
   resolveProviderInput,
   winnabilityClassSchema
-} from "./chunk-NG7OR6GO.js";
+} from "./chunk-ESH3LPYQ.js";
 
 // src/cli.ts
 import { pathToFileURL } from "url";

package/dist/index.js

@@ -1,11 +1,11 @@
 import {
   createServer
-} from "./chunk-FZAAGXTK.js";
+} from "./chunk-MRHYSEZB.js";
 import {
   loadConfig
-} from "./chunk-2TLRE2L3.js";
-import "./chunk-C6ZZH3XI.js";
-import "./chunk-NG7OR6GO.js";
+} from "./chunk-INXI6PO7.js";
+import "./chunk-G5UZTODF.js";
+import "./chunk-ESH3LPYQ.js";
 export {
   createServer,
   loadConfig

package/dist/{intelligence-service-ANRFEQWX.js → intelligence-service-AFAFGLCM.js}

@@ -1,7 +1,7 @@
 import {
   IntelligenceService
-} from "./chunk-C6ZZH3XI.js";
-import "./chunk-NG7OR6GO.js";
+} from "./chunk-G5UZTODF.js";
+import "./chunk-ESH3LPYQ.js";
 export {
   IntelligenceService
 };

package/dist/mcp.js

@@ -3,10 +3,10 @@ import {
   PACKAGE_VERSION,
   canonryMcpTools,
   createApiClient
-} from "./chunk-2TLRE2L3.js";
+} from "./chunk-INXI6PO7.js";
 import {
   isReadOnlyKey
-} from "./chunk-NG7OR6GO.js";
+} from "./chunk-ESH3LPYQ.js";
 
 // src/mcp/cli.ts
 import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js";

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@canonry/canonry",
-  "version": "4.162.0",
+  "version": "4.163.0",
   "type": "module",
   "description": "Agent-first open-source AEO operating platform - track how answer engines cite your domain",
   "license": "FSL-1.1-ALv2",
@@ -41,7 +41,7 @@
   },
   "dependencies": {
     "@anthropic-ai/sdk": "^0.91.1",
-    "@canonry/aeo-audit": "4.6.2",
+    "@canonry/aeo-audit": "4.7.0",
     "@fastify/rate-limit": "^11.2.0",
     "@fastify/static": "^9.1.1",
     "@google/genai": "^1.46.0",
@@ -71,28 +71,28 @@
     "tsup": "^8.5.1",
     "tsx": "^4.19.0",
     "@ainyc/canonry-api-client": "0.0.0",
-    "@ainyc/canonry-config": "0.0.0",
-    "@ainyc/canonry-integration-bing": "0.0.0",
+    "@ainyc/canonry-api-routes": "0.0.0",
     "@ainyc/canonry-contracts": "0.0.0",
     "@ainyc/canonry-db": "0.0.0",
+    "@ainyc/canonry-config": "0.0.0",
+    "@ainyc/canonry-integration-bing": "0.0.0",
     "@ainyc/canonry-integration-cloudflare-queue": "0.0.0",
     "@ainyc/canonry-integration-cloudflare-worker": "0.0.0",
-    "@ainyc/canonry-api-routes": "0.0.0",
-    "@ainyc/canonry-integration-cloud-run": "0.0.0",
     "@ainyc/canonry-integration-openai-ads": "0.0.0",
-    "@ainyc/canonry-integration-google": "0.0.0",
     "@ainyc/canonry-integration-commoncrawl": "0.0.0",
+    "@ainyc/canonry-integration-cloud-run": "0.0.0",
+    "@ainyc/canonry-integration-google": "0.0.0",
+    "@ainyc/canonry-integration-google-business-profile": "0.0.0",
     "@ainyc/canonry-integration-google-places": "0.0.0",
     "@ainyc/canonry-integration-traffic": "0.0.0",
-    "@ainyc/canonry-integration-google-business-profile": "0.0.0",
     "@ainyc/canonry-integration-wordpress": "0.0.0",
-    "@ainyc/canonry-provider-claude": "0.0.0",
     "@ainyc/canonry-provider-cdp": "0.0.0",
-    "@ainyc/canonry-provider-gemini": "0.0.0",
+    "@ainyc/canonry-provider-claude": "0.0.0",
     "@ainyc/canonry-intelligence": "0.0.0",
-    "@ainyc/canonry-provider-openai": "0.0.0",
+    "@ainyc/canonry-provider-gemini": "0.0.0",
     "@ainyc/canonry-provider-local": "0.0.0",
-    "@ainyc/canonry-provider-perplexity": "0.0.0"
+    "@ainyc/canonry-provider-perplexity": "0.0.0",
+    "@ainyc/canonry-provider-openai": "0.0.0"
   },
   "scripts": {
     "build": "tsx scripts/copy-agent-assets.ts && tsup && tsx build-web.ts",