@canonry/canonry 4.154.1 → 4.155.0

package/dist/{chunk-KRMLGFY3.js → chunk-VOL2BSM5.js}

@@ -403,6 +403,7 @@ import {
   normalizeProjectAliases,
   normalizeProjectDomain,
   normalizeQueryText,
+  normalizeSiteAuditRunRequest,
   normalizeUrlPath,
   normalizeUserName,
   notFound,
@@ -469,10 +470,17 @@ import {
   serializeRunError,
   settingsDtoSchema,
   siteAuditPagesResponseSchema,
+  siteAuditRequestIdentity,
   siteAuditRunRequestSchema,
   siteAuditRunResponseSchema,
   siteAuditScoreSchema,
   siteAuditTrendResponseSchema,
+  siteCrawlDeadLinksResponseSchema,
+  siteCrawlInternalLinksResponseSchema,
+  siteCrawlNeighborsResponseSchema,
+  siteCrawlPagesResponseSchema,
+  siteCrawlStructureResponseSchema,
+  siteCrawlSummarySchema,
   snapshotDiffResponseSchema,
   snapshotListResponseSchema,
   snapshotReportSchema,
@@ -521,10 +529,10 @@ import {
   wordpressSchemaDeployResultDtoSchema,
   wordpressSchemaStatusResultDtoSchema,
   wordpressStatusDtoSchema
-} from "./chunk-NNW7TJUZ.js";
+} from "./chunk-VIKPFI6F.js";
 
 // src/intelligence-service.ts
-import { eq as eq57, desc as desc26, asc as asc11, and as and45, ne as ne7, or as or11, inArray as inArray19, gte as gte13, lte as lte10 } from "drizzle-orm";
+import { eq as eq57, desc as desc26, asc as asc11, and as and45, ne as ne7, or as or12, inArray as inArray20, gte as gte13, lte as lte10 } from "drizzle-orm";
 
 // ../db/src/client.ts
 import { mkdirSync } from "fs";
@@ -615,6 +623,13 @@ __export(schema_exports, {
   schedules: () => schedules,
   siteAuditPages: () => siteAuditPages,
   siteAuditSnapshots: () => siteAuditSnapshots,
+  siteCrawlAttempts: () => siteCrawlAttempts,
+  siteCrawlEdges: () => siteCrawlEdges,
+  siteCrawlEventReceipts: () => siteCrawlEventReceipts,
+  siteCrawlFindings: () => siteCrawlFindings,
+  siteCrawlPages: () => siteCrawlPages,
+  siteCrawlRunRequests: () => siteCrawlRunRequests,
+  siteCrawlSnapshots: () => siteCrawlSnapshots,
   trafficSources: () => trafficSources,
   usageCounters: () => usageCounters,
   userSessions: () => userSessions,
@@ -839,6 +854,9 @@ var runs = sqliteTable("runs", {
   createdAt: text("created_at").notNull()
 }, (table) => [
   index("idx_runs_project").on(table.projectId),
+  // Enables child tables that carry project + run to enforce that the two IDs
+  // belong together, rather than trusting a caller-provided project ID.
+  uniqueIndex("idx_runs_project_id").on(table.projectId, table.id),
   index("idx_runs_status").on(table.status),
   index("idx_runs_source").on(table.sourceId),
   index("idx_runs_measurement_plan").on(table.projectId, table.measurementPlanVersionId, table.createdAt),
@@ -1185,6 +1203,213 @@ var siteAuditPages = sqliteTable("site_audit_pages", {
   index("idx_site_audit_pages_run").on(table.runId),
   index("idx_site_audit_pages_project_score").on(table.projectId, table.overallScore)
 ]);
+var siteCrawlRunRequests = sqliteTable("site_crawl_run_requests", {
+  runId: text("run_id").primaryKey(),
+  projectId: text("project_id").notNull().references(() => projects.id, { onDelete: "cascade" }),
+  identityKey: text("identity_key").notNull(),
+  effectiveOptions: text("effective_options", { mode: "json" }).$type().notNull(),
+  createdAt: text("created_at").notNull()
+}, (table) => [
+  index("idx_site_crawl_run_requests_project").on(table.projectId, table.runId),
+  foreignKey({
+    name: "site_crawl_run_requests_project_run_fk",
+    columns: [table.projectId, table.runId],
+    foreignColumns: [runs.projectId, runs.id]
+  }).onDelete("cascade")
+]);
+var siteCrawlAttempts = sqliteTable("site_crawl_attempts", {
+  id: text("id").primaryKey(),
+  projectId: text("project_id").notNull().references(() => projects.id, { onDelete: "cascade" }),
+  runId: text("run_id").notNull(),
+  attemptNumber: integer("attempt_number").notNull(),
+  state: text("state").notNull().default("queued"),
+  lastEventSequence: integer("last_event_sequence").notNull().default(0),
+  lastEventChecksum: text("last_event_checksum"),
+  pagesDiscovered: integer("pages_discovered").notNull().default(0),
+  pagesFetched: integer("pages_fetched").notNull().default(0),
+  pagesEligible: integer("pages_eligible").notNull().default(0),
+  pagesErrored: integer("pages_errored").notNull().default(0),
+  edgesDiscovered: integer("edges_discovered").notNull().default(0),
+  startedAt: text("started_at"),
+  finishedAt: text("finished_at"),
+  error: text("error"),
+  createdAt: text("created_at").notNull(),
+  updatedAt: text("updated_at").notNull()
+}, (table) => [
+  uniqueIndex("idx_site_crawl_attempts_run_number").on(table.runId, table.attemptNumber),
+  uniqueIndex("idx_site_crawl_attempts_project_run_id").on(table.projectId, table.runId, table.id),
+  index("idx_site_crawl_attempts_project_run").on(table.projectId, table.runId),
+  foreignKey({
+    name: "site_crawl_attempts_project_run_fk",
+    columns: [table.projectId, table.runId],
+    foreignColumns: [runs.projectId, runs.id]
+  }).onDelete("cascade")
+]);
+var siteCrawlSnapshots = sqliteTable("site_crawl_snapshots", {
+  id: text("id").primaryKey(),
+  projectId: text("project_id").notNull().references(() => projects.id, { onDelete: "cascade" }),
+  runId: text("run_id").notNull(),
+  attemptId: text("attempt_id"),
+  rootUrl: text("root_url").notNull(),
+  crawlSchemaVersion: text("crawl_schema_version").notNull().default("1.0"),
+  engineVersion: text("engine_version").notNull().default(""),
+  normalizationVersion: text("normalization_version").notNull().default(""),
+  indexabilityVersion: text("indexability_version").notNull().default(""),
+  linkScoreVersion: text("link_score_version").notNull().default(""),
+  effectiveOptions: text("effective_options", { mode: "json" }).$type().notNull().default({}),
+  pageBudget: integer("page_budget"),
+  edgeBudget: integer("edge_budget"),
+  maxDepth: integer("max_depth"),
+  checkDeadLinks: integer("check_dead_links", { mode: "boolean" }).notNull().default(false),
+  complete: integer("complete", { mode: "boolean" }).notNull().default(false),
+  termination: text("termination").notNull().default("unknown"),
+  detailsAvailable: integer("details_available", { mode: "boolean" }).notNull().default(false),
+  pagesDiscovered: integer("pages_discovered").notNull().default(0),
+  pagesFetched: integer("pages_fetched").notNull().default(0),
+  pagesEligible: integer("pages_eligible").notNull().default(0),
+  pagesErrored: integer("pages_errored").notNull().default(0),
+  edgesDiscovered: integer("edges_discovered").notNull().default(0),
+  findingsCount: integer("findings_count").notNull().default(0),
+  deadLinkState: text("dead_link_state").notNull().default("disabled"),
+  deadLinksChecked: integer("dead_links_checked").notNull().default(0),
+  deadLinksFound: integer("dead_links_found").notNull().default(0),
+  createdAt: text("created_at").notNull(),
+  updatedAt: text("updated_at").notNull()
+}, (table) => [
+  uniqueIndex("idx_site_crawl_snapshots_run").on(table.runId),
+  index("idx_site_crawl_snapshots_project_created").on(table.projectId, table.createdAt),
+  foreignKey({
+    name: "site_crawl_snapshots_project_run_fk",
+    columns: [table.projectId, table.runId],
+    foreignColumns: [runs.projectId, runs.id]
+  }).onDelete("cascade"),
+  foreignKey({
+    name: "site_crawl_snapshots_attempt_fk",
+    columns: [table.projectId, table.runId, table.attemptId],
+    foreignColumns: [siteCrawlAttempts.projectId, siteCrawlAttempts.runId, siteCrawlAttempts.id]
+  })
+]);
+var siteCrawlPages = sqliteTable("site_crawl_pages", {
+  id: text("id").primaryKey(),
+  projectId: text("project_id").notNull().references(() => projects.id, { onDelete: "cascade" }),
+  runId: text("run_id").notNull(),
+  attemptId: text("attempt_id").notNull(),
+  nodeKey: text("node_key").notNull(),
+  url: text("url").notNull(),
+  path: text("path").notNull(),
+  parentPath: text("parent_path").notNull(),
+  discoverySource: text("discovery_source").notNull().default("crawl"),
+  discoveryProvenance: text("discovery_provenance", { mode: "json" }).$type().notNull().default([]),
+  sitemapMetadata: text("sitemap_metadata", { mode: "json" }).$type().notNull().default({}),
+  fetchState: text("fetch_state").notNull().default("queued"),
+  fetchedAt: text("fetched_at"),
+  httpStatus: integer("http_status"),
+  contentType: text("content_type"),
+  finalUrl: text("final_url"),
+  redirectChain: text("redirect_chain", { mode: "json" }).$type().notNull().default([]),
+  directives: text("directives", { mode: "json" }).$type().notNull().default({}),
+  canonicalUrl: text("canonical_url"),
+  canonicalNodeKey: text("canonical_node_key"),
+  indexabilityState: text("indexability_state").notNull().default("unknown"),
+  indexabilityReasons: text("indexability_reasons", { mode: "json" }).$type().notNull().default([]),
+  auditState: text("audit_state").notNull().default("pending"),
+  auditScore: real("audit_score"),
+  auditFields: text("audit_fields", { mode: "json" }).$type().notNull().default({}),
+  inventoryEligible: integer("inventory_eligible", { mode: "boolean" }).notNull().default(false),
+  depth: integer("depth"),
+  inboundUniqueEdges: integer("inbound_unique_edges").notNull().default(0),
+  outboundUniqueEdges: integer("outbound_unique_edges").notNull().default(0),
+  inboundOccurrences: integer("inbound_occurrences").notNull().default(0),
+  outboundOccurrences: integer("outbound_occurrences").notNull().default(0),
+  linkScoreRaw: real("link_score_raw"),
+  linkScoreNormalized: real("link_score_normalized"),
+  createdAt: text("created_at").notNull(),
+  updatedAt: text("updated_at").notNull()
+}, (table) => [
+  uniqueIndex("idx_site_crawl_pages_attempt_node").on(table.projectId, table.runId, table.attemptId, table.nodeKey),
+  index("idx_site_crawl_pages_read").on(table.projectId, table.runId, table.attemptId, table.inventoryEligible, table.auditScore, table.url),
+  index("idx_site_crawl_pages_parent").on(table.projectId, table.runId, table.attemptId, table.parentPath, table.path),
+  index("idx_site_crawl_pages_url").on(table.projectId, table.runId, table.attemptId, table.url),
+  foreignKey({
+    name: "site_crawl_pages_attempt_fk",
+    columns: [table.projectId, table.runId, table.attemptId],
+    foreignColumns: [siteCrawlAttempts.projectId, siteCrawlAttempts.runId, siteCrawlAttempts.id]
+  }).onDelete("cascade")
+]);
+var siteCrawlEdges = sqliteTable("site_crawl_edges", {
+  id: text("id").primaryKey(),
+  projectId: text("project_id").notNull().references(() => projects.id, { onDelete: "cascade" }),
+  runId: text("run_id").notNull(),
+  attemptId: text("attempt_id").notNull(),
+  edgeKey: text("edge_key").notNull(),
+  sourceNodeKey: text("source_node_key").notNull(),
+  sourceUrl: text("source_url").notNull(),
+  targetNodeKey: text("target_node_key"),
+  targetUrl: text("target_url").notNull(),
+  relation: text("relation").notNull().default("link"),
+  internal: integer("internal", { mode: "boolean" }).notNull().default(true),
+  followable: integer("followable", { mode: "boolean" }).notNull().default(true),
+  occurrences: integer("occurrences").notNull().default(1),
+  followableOccurrences: integer("followable_occurrences").notNull().default(1),
+  nofollowOccurrences: integer("nofollow_occurrences").notNull().default(0),
+  anchors: text("anchors", { mode: "json" }).$type().notNull().default([]),
+  createdAt: text("created_at").notNull(),
+  updatedAt: text("updated_at").notNull()
+}, (table) => [
+  uniqueIndex("idx_site_crawl_edges_attempt_key").on(table.projectId, table.runId, table.attemptId, table.edgeKey),
+  index("idx_site_crawl_edges_outbound").on(table.projectId, table.runId, table.attemptId, table.sourceNodeKey, table.edgeKey),
+  index("idx_site_crawl_edges_inbound").on(table.projectId, table.runId, table.attemptId, table.targetNodeKey, table.edgeKey),
+  index("idx_site_crawl_edges_source_url").on(table.projectId, table.runId, table.attemptId, table.sourceUrl),
+  index("idx_site_crawl_edges_target_url").on(table.projectId, table.runId, table.attemptId, table.targetUrl),
+  foreignKey({
+    name: "site_crawl_edges_attempt_fk",
+    columns: [table.projectId, table.runId, table.attemptId],
+    foreignColumns: [siteCrawlAttempts.projectId, siteCrawlAttempts.runId, siteCrawlAttempts.id]
+  }).onDelete("cascade")
+]);
+var siteCrawlFindings = sqliteTable("site_crawl_findings", {
+  id: text("id").primaryKey(),
+  projectId: text("project_id").notNull().references(() => projects.id, { onDelete: "cascade" }),
+  runId: text("run_id").notNull(),
+  attemptId: text("attempt_id").notNull(),
+  findingKey: text("finding_key").notNull(),
+  findingType: text("finding_type").notNull(),
+  severity: text("severity").notNull().default("info"),
+  sourceNodeKey: text("source_node_key"),
+  sourceUrl: text("source_url"),
+  targetNodeKey: text("target_node_key"),
+  targetUrl: text("target_url"),
+  evidence: text("evidence", { mode: "json" }).$type().notNull().default({}),
+  createdAt: text("created_at").notNull(),
+  updatedAt: text("updated_at").notNull()
+}, (table) => [
+  uniqueIndex("idx_site_crawl_findings_attempt_key").on(table.projectId, table.runId, table.attemptId, table.findingKey),
+  index("idx_site_crawl_findings_type").on(table.projectId, table.runId, table.attemptId, table.findingType, table.findingKey),
+  foreignKey({
+    name: "site_crawl_findings_attempt_fk",
+    columns: [table.projectId, table.runId, table.attemptId],
+    foreignColumns: [siteCrawlAttempts.projectId, siteCrawlAttempts.runId, siteCrawlAttempts.id]
+  }).onDelete("cascade")
+]);
+var siteCrawlEventReceipts = sqliteTable("site_crawl_event_receipts", {
+  id: text("id").primaryKey(),
+  projectId: text("project_id").notNull().references(() => projects.id, { onDelete: "cascade" }),
+  runId: text("run_id").notNull(),
+  attemptId: text("attempt_id").notNull(),
+  sequence: integer("sequence").notNull(),
+  batchId: text("batch_id").notNull(),
+  checksum: text("checksum").notNull(),
+  receipt: text("receipt", { mode: "json" }).$type().notNull().default({}),
+  createdAt: text("created_at").notNull()
+}, (table) => [
+  uniqueIndex("idx_site_crawl_receipts_attempt_event").on(table.attemptId, table.sequence, table.batchId),
+  index("idx_site_crawl_receipts_project_run").on(table.projectId, table.runId, table.attemptId),
+  foreignKey({
+    name: "site_crawl_receipts_attempt_fk",
+    columns: [table.projectId, table.runId, table.attemptId],
+    foreignColumns: [siteCrawlAttempts.projectId, siteCrawlAttempts.runId, siteCrawlAttempts.id]
+  }).onDelete("cascade")
+]);
 var bingCoverageSnapshots = sqliteTable("bing_coverage_snapshots", {
   id: text("id").primaryKey(),
   projectId: text("project_id").notNull().references(() => projects.id, { onDelete: "cascade" }),
@@ -5096,6 +5321,211 @@ var MIGRATION_VERSIONS = [
       `ALTER TABLE gsc_coverage_snapshots ADD COLUMN verified_by_inspection INTEGER NOT NULL DEFAULT 0`,
       `ALTER TABLE gsc_coverage_snapshots ADD COLUMN derived_from_impressions INTEGER NOT NULL DEFAULT 0`
     ]
+  },
+  {
+    // Crawl persistence is intentionally separate from the legacy
+    // site_audit_snapshots/site_audit_pages scorecard model. Older engines can
+    // keep reading and writing that legacy surface after a rollback; new tables
+    // only enrich an audit run when the crawler writes them.
+    version: 126,
+    name: "site-crawl-persistence",
+    statements: [
+      // Composite child FKs below enforce that project_id and run_id are a
+      // matched pair. `runs.id` is globally unique already; this named unique
+      // index provides SQLite's required parent key for the composite check.
+      `CREATE UNIQUE INDEX IF NOT EXISTS idx_runs_project_id ON runs(project_id, id)`,
+      `CREATE TABLE IF NOT EXISTS site_crawl_run_requests (
+        run_id           TEXT PRIMARY KEY,
+        project_id       TEXT NOT NULL REFERENCES projects(id) ON DELETE CASCADE,
+        identity_key     TEXT NOT NULL,
+        effective_options TEXT NOT NULL,
+        created_at       TEXT NOT NULL,
+        FOREIGN KEY (project_id, run_id) REFERENCES runs(project_id, id) ON DELETE CASCADE
+      )`,
+      `CREATE INDEX IF NOT EXISTS idx_site_crawl_run_requests_project
+        ON site_crawl_run_requests(project_id, run_id)`,
+      `CREATE TABLE IF NOT EXISTS site_crawl_attempts (
+        id                  TEXT PRIMARY KEY,
+        project_id          TEXT NOT NULL REFERENCES projects(id) ON DELETE CASCADE,
+        run_id              TEXT NOT NULL,
+        attempt_number      INTEGER NOT NULL,
+        state               TEXT NOT NULL DEFAULT 'queued',
+        last_event_sequence INTEGER NOT NULL DEFAULT 0,
+        last_event_checksum TEXT,
+        pages_discovered    INTEGER NOT NULL DEFAULT 0,
+        pages_fetched       INTEGER NOT NULL DEFAULT 0,
+        pages_eligible      INTEGER NOT NULL DEFAULT 0,
+        pages_errored       INTEGER NOT NULL DEFAULT 0,
+        edges_discovered    INTEGER NOT NULL DEFAULT 0,
+        started_at          TEXT,
+        finished_at         TEXT,
+        error               TEXT,
+        created_at          TEXT NOT NULL,
+        updated_at          TEXT NOT NULL,
+        FOREIGN KEY (project_id, run_id) REFERENCES runs(project_id, id) ON DELETE CASCADE
+      )`,
+      `CREATE UNIQUE INDEX IF NOT EXISTS idx_site_crawl_attempts_run_number
+        ON site_crawl_attempts(run_id, attempt_number)`,
+      `CREATE UNIQUE INDEX IF NOT EXISTS idx_site_crawl_attempts_project_run_id
+        ON site_crawl_attempts(project_id, run_id, id)`,
+      `CREATE INDEX IF NOT EXISTS idx_site_crawl_attempts_project_run
+        ON site_crawl_attempts(project_id, run_id)`,
+      `CREATE TABLE IF NOT EXISTS site_crawl_snapshots (
+        id                    TEXT PRIMARY KEY,
+        project_id            TEXT NOT NULL REFERENCES projects(id) ON DELETE CASCADE,
+        run_id                TEXT NOT NULL,
+        attempt_id            TEXT,
+        root_url              TEXT NOT NULL,
+        crawl_schema_version  TEXT NOT NULL DEFAULT '1.0',
+        engine_version        TEXT NOT NULL DEFAULT '',
+        normalization_version TEXT NOT NULL DEFAULT '',
+        indexability_version  TEXT NOT NULL DEFAULT '',
+        link_score_version    TEXT NOT NULL DEFAULT '',
+        effective_options     TEXT NOT NULL DEFAULT '{}',
+        page_budget           INTEGER,
+        edge_budget           INTEGER,
+        max_depth             INTEGER,
+        check_dead_links      INTEGER NOT NULL DEFAULT 0,
+        complete              INTEGER NOT NULL DEFAULT 0,
+        termination           TEXT NOT NULL DEFAULT 'unknown',
+        details_available     INTEGER NOT NULL DEFAULT 0,
+        pages_discovered      INTEGER NOT NULL DEFAULT 0,
+        pages_fetched         INTEGER NOT NULL DEFAULT 0,
+        pages_eligible        INTEGER NOT NULL DEFAULT 0,
+        pages_errored         INTEGER NOT NULL DEFAULT 0,
+        edges_discovered      INTEGER NOT NULL DEFAULT 0,
+        findings_count        INTEGER NOT NULL DEFAULT 0,
+        dead_link_state       TEXT NOT NULL DEFAULT 'disabled',
+        dead_links_checked    INTEGER NOT NULL DEFAULT 0,
+        dead_links_found      INTEGER NOT NULL DEFAULT 0,
+        created_at            TEXT NOT NULL,
+        updated_at            TEXT NOT NULL,
+        FOREIGN KEY (project_id, run_id) REFERENCES runs(project_id, id) ON DELETE CASCADE,
+        FOREIGN KEY (project_id, run_id, attempt_id)
+          REFERENCES site_crawl_attempts(project_id, run_id, id)
+      )`,
+      `CREATE UNIQUE INDEX IF NOT EXISTS idx_site_crawl_snapshots_run ON site_crawl_snapshots(run_id)`,
+      `CREATE INDEX IF NOT EXISTS idx_site_crawl_snapshots_project_created
+        ON site_crawl_snapshots(project_id, created_at)`,
+      `CREATE TABLE IF NOT EXISTS site_crawl_pages (
+        id                   TEXT PRIMARY KEY,
+        project_id           TEXT NOT NULL REFERENCES projects(id) ON DELETE CASCADE,
+        run_id               TEXT NOT NULL,
+        attempt_id           TEXT NOT NULL,
+        node_key             TEXT NOT NULL,
+        url                  TEXT NOT NULL,
+        path                 TEXT NOT NULL,
+        parent_path          TEXT NOT NULL,
+        discovery_source     TEXT NOT NULL DEFAULT 'crawl',
+        discovery_provenance TEXT NOT NULL DEFAULT '[]',
+        sitemap_metadata     TEXT NOT NULL DEFAULT '{}',
+        fetch_state          TEXT NOT NULL DEFAULT 'queued',
+        fetched_at           TEXT,
+        http_status          INTEGER,
+        content_type         TEXT,
+        final_url            TEXT,
+        redirect_chain       TEXT NOT NULL DEFAULT '[]',
+        directives           TEXT NOT NULL DEFAULT '{}',
+        canonical_url        TEXT,
+        canonical_node_key   TEXT,
+        indexability_state   TEXT NOT NULL DEFAULT 'unknown',
+        indexability_reasons TEXT NOT NULL DEFAULT '[]',
+        audit_state          TEXT NOT NULL DEFAULT 'pending',
+        audit_score          REAL,
+        audit_fields         TEXT NOT NULL DEFAULT '{}',
+        inventory_eligible   INTEGER NOT NULL DEFAULT 0,
+        depth                INTEGER,
+        inbound_unique_edges INTEGER NOT NULL DEFAULT 0,
+        outbound_unique_edges INTEGER NOT NULL DEFAULT 0,
+        inbound_occurrences  INTEGER NOT NULL DEFAULT 0,
+        outbound_occurrences INTEGER NOT NULL DEFAULT 0,
+        link_score_raw       REAL,
+        link_score_normalized REAL,
+        created_at           TEXT NOT NULL,
+        updated_at           TEXT NOT NULL,
+        FOREIGN KEY (project_id, run_id, attempt_id)
+          REFERENCES site_crawl_attempts(project_id, run_id, id) ON DELETE CASCADE
+      )`,
+      `CREATE UNIQUE INDEX IF NOT EXISTS idx_site_crawl_pages_attempt_node
+        ON site_crawl_pages(project_id, run_id, attempt_id, node_key)`,
+      `CREATE INDEX IF NOT EXISTS idx_site_crawl_pages_read
+        ON site_crawl_pages(project_id, run_id, attempt_id, inventory_eligible, audit_score, url)`,
+      `CREATE INDEX IF NOT EXISTS idx_site_crawl_pages_parent
+        ON site_crawl_pages(project_id, run_id, attempt_id, parent_path, path)`,
+      `CREATE INDEX IF NOT EXISTS idx_site_crawl_pages_url
+        ON site_crawl_pages(project_id, run_id, attempt_id, url)`,
+      `CREATE TABLE IF NOT EXISTS site_crawl_edges (
+        id              TEXT PRIMARY KEY,
+        project_id      TEXT NOT NULL REFERENCES projects(id) ON DELETE CASCADE,
+        run_id          TEXT NOT NULL,
+        attempt_id      TEXT NOT NULL,
+        edge_key        TEXT NOT NULL,
+        source_node_key TEXT NOT NULL,
+        source_url      TEXT NOT NULL,
+        target_node_key TEXT,
+        target_url      TEXT NOT NULL,
+        relation        TEXT NOT NULL DEFAULT 'link',
+        internal        INTEGER NOT NULL DEFAULT 1,
+        followable      INTEGER NOT NULL DEFAULT 1,
+        occurrences     INTEGER NOT NULL DEFAULT 1,
+        followable_occurrences INTEGER NOT NULL DEFAULT 1,
+        nofollow_occurrences   INTEGER NOT NULL DEFAULT 0,
+        anchors         TEXT NOT NULL DEFAULT '[]',
+        created_at      TEXT NOT NULL,
+        updated_at      TEXT NOT NULL,
+        FOREIGN KEY (project_id, run_id, attempt_id)
+          REFERENCES site_crawl_attempts(project_id, run_id, id) ON DELETE CASCADE
+      )`,
+      `CREATE UNIQUE INDEX IF NOT EXISTS idx_site_crawl_edges_attempt_key
+        ON site_crawl_edges(project_id, run_id, attempt_id, edge_key)`,
+      `CREATE INDEX IF NOT EXISTS idx_site_crawl_edges_outbound
+        ON site_crawl_edges(project_id, run_id, attempt_id, source_node_key, edge_key)`,
+      `CREATE INDEX IF NOT EXISTS idx_site_crawl_edges_inbound
+        ON site_crawl_edges(project_id, run_id, attempt_id, target_node_key, edge_key)`,
+      `CREATE INDEX IF NOT EXISTS idx_site_crawl_edges_source_url
+        ON site_crawl_edges(project_id, run_id, attempt_id, source_url)`,
+      `CREATE INDEX IF NOT EXISTS idx_site_crawl_edges_target_url
+        ON site_crawl_edges(project_id, run_id, attempt_id, target_url)`,
+      `CREATE TABLE IF NOT EXISTS site_crawl_findings (
+        id              TEXT PRIMARY KEY,
+        project_id      TEXT NOT NULL REFERENCES projects(id) ON DELETE CASCADE,
+        run_id          TEXT NOT NULL,
+        attempt_id      TEXT NOT NULL,
+        finding_key     TEXT NOT NULL,
+        finding_type    TEXT NOT NULL,
+        severity        TEXT NOT NULL DEFAULT 'info',
+        source_node_key TEXT,
+        source_url      TEXT,
+        target_node_key TEXT,
+        target_url      TEXT,
+        evidence        TEXT NOT NULL DEFAULT '{}',
+        created_at      TEXT NOT NULL,
+        updated_at      TEXT NOT NULL,
+        FOREIGN KEY (project_id, run_id, attempt_id)
+          REFERENCES site_crawl_attempts(project_id, run_id, id) ON DELETE CASCADE
+      )`,
+      `CREATE UNIQUE INDEX IF NOT EXISTS idx_site_crawl_findings_attempt_key
+        ON site_crawl_findings(project_id, run_id, attempt_id, finding_key)`,
+      `CREATE INDEX IF NOT EXISTS idx_site_crawl_findings_type
+        ON site_crawl_findings(project_id, run_id, attempt_id, finding_type, finding_key)`,
+      `CREATE TABLE IF NOT EXISTS site_crawl_event_receipts (
+        id         TEXT PRIMARY KEY,
+        project_id TEXT NOT NULL REFERENCES projects(id) ON DELETE CASCADE,
+        run_id     TEXT NOT NULL,
+        attempt_id TEXT NOT NULL,
+        sequence   INTEGER NOT NULL,
+        batch_id   TEXT NOT NULL,
+        checksum   TEXT NOT NULL,
+        receipt    TEXT NOT NULL DEFAULT '{}',
+        created_at TEXT NOT NULL,
+        FOREIGN KEY (project_id, run_id, attempt_id)
+          REFERENCES site_crawl_attempts(project_id, run_id, id) ON DELETE CASCADE
+      )`,
+      `CREATE UNIQUE INDEX IF NOT EXISTS idx_site_crawl_receipts_attempt_event
+        ON site_crawl_event_receipts(attempt_id, sequence, batch_id)`,
+      `CREATE INDEX IF NOT EXISTS idx_site_crawl_receipts_project_run
+        ON site_crawl_event_receipts(project_id, run_id, attempt_id)`
+    ]
   }
 ];
 function addRunsMeasurementPlanVersionForeignKey(tx) {
@@ -7968,7 +8398,7 @@ function isOverviewRead(url) {
 }
 function isTechnicalAeoRead(url) {
   const rest = projectRouteRest(url);
-  return rest === "technical-aeo" || rest === "technical-aeo/pages" || rest === "technical-aeo/trend";
+  return rest === "technical-aeo" || rest === "technical-aeo/pages" || rest === "technical-aeo/trend" || rest === "technical-aeo/crawl" || rest === "technical-aeo/crawl/pages" || rest === "technical-aeo/structure" || rest === "technical-aeo/internal-links" || rest === "technical-aeo/internal-links/neighbors" || rest === "technical-aeo/dead-links";
 }
 function isReportRead(url) {
   const rest = projectRouteRest(url);
@@ -9206,7 +9636,7 @@ function parseCompetitorBatch(value) {
 
 // ../api-routes/src/runs.ts
 import crypto12 from "crypto";
-import { and as and6, eq as eq11, asc, desc as desc3, or as or2, sql as sql6 } from "drizzle-orm";
+import { and as and6, eq as eq11, asc, desc as desc3, inArray as inArray4, or as or2, sql as sql6 } from "drizzle-orm";
 import { gte as gte2 } from "drizzle-orm";
 
 // ../api-routes/src/run-queue.ts
@@ -10954,7 +11384,12 @@ async function runRoutes(app, opts) {
     const terminalStatuses = /* @__PURE__ */ new Set(["completed", "partial", "failed", "cancelled"]);
     if (terminalStatuses.has(run.status)) throw runNotCancellable(run.id, run.status);
     const now = (/* @__PURE__ */ new Date()).toISOString();
-    app.db.update(runs).set({ status: "cancelled", finishedAt: now, error: serializeRunError({ message: "Cancelled by user" }) }).where(eq11(runs.id, run.id)).run();
+    const cancelled = app.db.update(runs).set({ status: "cancelled", finishedAt: now, error: serializeRunError({ message: "Cancelled by user" }) }).where(and6(eq11(runs.id, run.id), inArray4(runs.status, ["queued", "running"]))).run();
+    if (cancelled.changes === 0) {
+      const current = app.db.select().from(runs).where(eq11(runs.id, run.id)).get();
+      if (!current) throw notFound("Run", run.id);
+      throw runNotCancellable(run.id, current.status);
+    }
     writeAuditLog(app.db, {
       projectId: run.projectId,
       actor: "api",
@@ -10962,6 +11397,7 @@ async function runRoutes(app, opts) {
       entityType: "run",
       entityId: run.id
     });
+    opts.onRunCancelled?.(run.id, run.projectId);
     const updated = app.db.select().from(runs).where(eq11(runs.id, run.id)).get();
     return reply.send(formatRun(updated));
   });
@@ -12231,7 +12667,7 @@ async function measurementServiceRoutes(app, opts) {
 
 // ../api-routes/src/measurement-draft.ts
 import crypto17 from "crypto";
-import { and as and9, asc as asc2, desc as desc5, eq as eq14, inArray as inArray4 } from "drizzle-orm";
+import { and as and9, asc as asc2, desc as desc5, eq as eq14, inArray as inArray5 } from "drizzle-orm";
 
 // ../api-routes/src/measurement-draft-repo.ts
 import crypto15 from "crypto";
@@ -14753,7 +15189,7 @@ async function measurementDraftRoutes(app, opts = {}) {
     const sets = app.db.select().from(measurementQuerySets).where(eq14(measurementQuerySets.projectId, project.id)).orderBy(asc2(measurementQuerySets.name)).all();
     const counts2 = /* @__PURE__ */ new Map();
     if (sets.length) {
-      for (const item2 of app.db.select({ querySetId: measurementQuerySetItems.querySetId }).from(measurementQuerySetItems).where(inArray4(measurementQuerySetItems.querySetId, sets.map((set) => set.id))).all()) {
+      for (const item2 of app.db.select({ querySetId: measurementQuerySetItems.querySetId }).from(measurementQuerySetItems).where(inArray5(measurementQuerySetItems.querySetId, sets.map((set) => set.id))).all()) {
         counts2.set(item2.querySetId, (counts2.get(item2.querySetId) ?? 0) + 1);
       }
     }
@@ -16276,10 +16712,10 @@ async function measurementPropertyEvidenceRoutes(app) {
 }
 
 // ../api-routes/src/measurement-portfolio-reads.ts
-import { and as and14, desc as desc6, eq as eq20, inArray as inArray6, lt as lt3, or as or3 } from "drizzle-orm";
+import { and as and14, desc as desc6, eq as eq20, inArray as inArray7, lt as lt3, or as or3 } from "drizzle-orm";
 
 // ../api-routes/src/measurement-question-reads.ts
-import { and as and13, eq as eq19, getTableColumns, inArray as inArray5, sql as sql7 } from "drizzle-orm";
+import { and as and13, eq as eq19, getTableColumns, inArray as inArray6, sql as sql7 } from "drizzle-orm";
 var DEFAULT_LIMIT2 = 50;
 var TERMINAL_STATUSES = /* @__PURE__ */ new Set([
   RunStatuses.completed,
@@ -16395,7 +16831,7 @@ function selectedQuestionSnapshots(db, runId, manifest, scope) {
   if (scope !== void 0) {
     const executionIds = [...new Set(manifest.expectedSlots.map((slot) => slot.executionId))];
     if (executionIds.length === 0) return [];
-    conditions.push(inArray5(querySnapshots.measurementExecutionId, executionIds));
+    conditions.push(inArray6(querySnapshots.measurementExecutionId, executionIds));
   }
   if (scope?.provider !== void 0) {
     conditions.push(sql7`lower(trim(${querySnapshots.provider})) = ${normalizeText(scope.provider)}`);
@@ -16410,7 +16846,7 @@ function rawResponsesForLegacyQuestionSnapshots(db, runId, snapshots) {
     rawResponse: querySnapshots.rawResponse
   }).from(querySnapshots).where(and13(
     eq19(querySnapshots.runId, runId),
-    inArray5(querySnapshots.id, legacyIds)
+    inArray6(querySnapshots.id, legacyIds)
   )).all().map((snapshot) => [snapshot.id, snapshot.rawResponse]));
 }
 function withLegacyRawEvidence(db, runId, snapshots) {
@@ -17056,7 +17492,7 @@ function previousComparableRun(db, active, current) {
     eq20(runs.projectId, current.projectId),
     eq20(runs.measurementPlanVersionId, active.version.id),
     eq20(runs.kind, RunKinds["answer-visibility"]),
-    inArray6(runs.status, [RunStatuses.completed, RunStatuses.partial]),
+    inArray7(runs.status, [RunStatuses.completed, RunStatuses.partial]),
     or3(
       lt3(runs.createdAt, current.createdAt),
       and14(eq20(runs.createdAt, current.createdAt), lt3(runs.id, current.id))
@@ -17882,7 +18318,7 @@ function normalizeCompetitorList2(domains) {
 }
 
 // ../api-routes/src/history.ts
-import { and as and16, asc as asc3, desc as desc7, eq as eq22, gte as gte3, inArray as inArray7 } from "drizzle-orm";
+import { and as and16, asc as asc3, desc as desc7, eq as eq22, gte as gte3, inArray as inArray8 } from "drizzle-orm";
 
 // ../api-routes/src/notification-redaction.ts
 var REDACTED_URL = {
@@ -17968,7 +18404,7 @@ async function historyRoutes(app) {
       recommendedCompetitors: querySnapshots.recommendedCompetitors,
       location: querySnapshots.location,
       createdAt: querySnapshots.createdAt
-    }).from(querySnapshots).leftJoin(queries, eq22(querySnapshots.queryId, queries.id)).where(inArray7(querySnapshots.runId, projectRuns.map((r) => r.id))).orderBy(desc7(querySnapshots.createdAt)).all();
+    }).from(querySnapshots).leftJoin(queries, eq22(querySnapshots.queryId, queries.id)).where(inArray8(querySnapshots.runId, projectRuns.map((r) => r.id))).orderBy(desc7(querySnapshots.createdAt)).all();
     const locationFilter = request.query.location;
     const filtered = locationFilter !== void 0 ? allSnapshots.filter((s) => s.location === (locationFilter || null)) : allSnapshots;
     const total = filtered.length;
@@ -18014,7 +18450,7 @@ async function historyRoutes(app) {
       return reply.send([]);
     }
     const runIds = new Set(projectRuns.map((r) => r.id));
-    const rawSnapshots = app.db.select().from(querySnapshots).where(inArray7(querySnapshots.runId, [...runIds])).all();
+    const rawSnapshots = app.db.select().from(querySnapshots).where(inArray8(querySnapshots.runId, [...runIds])).all();
     const timelineLocationFilter = request.query.location;
     const filteredSnapshots = timelineLocationFilter !== void 0 ? rawSnapshots.filter((s) => s.location === (timelineLocationFilter || null)) : rawSnapshots;
     const allSnapshots = filteredSnapshots.map((snapshot) => ({
@@ -18135,7 +18571,7 @@ async function historyRoutes(app) {
       throw validationError("Both run1 and run2 query params are required");
     }
     const requestedRunIds = [.../* @__PURE__ */ new Set([run1, run2])];
-    const runRows = app.db.select({ id: runs.id, projectId: runs.projectId }).from(runs).where(inArray7(runs.id, requestedRunIds)).all();
+    const runRows = app.db.select({ id: runs.id, projectId: runs.projectId }).from(runs).where(inArray8(runs.id, requestedRunIds)).all();
     const runsById = new Map(runRows.map((row) => [row.id, row]));
     for (const runId of requestedRunIds) {
       const run = runsById.get(runId);
@@ -18233,7 +18669,7 @@ function parseOptionalPositiveInt(value, max) {
 }
 
 // ../api-routes/src/analytics.ts
-import { and as and17, desc as desc8, eq as eq23, gte as gte4, inArray as inArray8, lt as lt4 } from "drizzle-orm";
+import { and as and17, desc as desc8, eq as eq23, gte as gte4, inArray as inArray9, lt as lt4 } from "drizzle-orm";
 
 // ../api-routes/src/model-evidence.ts
 function classifyModelEvidence(values) {
@@ -18368,7 +18804,7 @@ async function analyticsRoutes(app) {
     const projectRuns = app.db.select().from(runs).where(and17(
       eq23(runs.projectId, project.id),
       eq23(runs.kind, RunKinds["answer-visibility"]),
-      inArray8(runs.status, [RunStatuses.completed, RunStatuses.partial]),
+      inArray9(runs.status, [RunStatuses.completed, RunStatuses.partial]),
       notProbeRun(),
       cutoff ? gte4(runs.createdAt, cutoff) : void 0
     )).orderBy(desc8(runs.createdAt)).all().filter((r) => measuredWhole(app.db, r)).sort((a, b) => a.createdAt.localeCompare(b.createdAt));
@@ -18411,7 +18847,7 @@ async function analyticsRoutes(app) {
       citationState: querySnapshots.citationState,
       answerMentioned: querySnapshots.answerMentioned,
       answerText: querySnapshots.answerText
-    }).from(querySnapshots).where(inArray8(querySnapshots.runId, runIds)).all();
+    }).from(querySnapshots).where(inArray9(querySnapshots.runId, runIds)).all();
     const rawSnapshots = referenceBasket ? loadedSnapshots.filter(
       (s) => s.queryId !== null || s.queryText !== null && referenceBasket.has(normalizeQueryText(s.queryText))
     ) : filterTrackedSnapshots(loadedSnapshots);
@@ -18453,7 +18889,7 @@ async function analyticsRoutes(app) {
       const anchorRunPredicate = and17(
         eq23(runs.projectId, project.id),
         eq23(runs.kind, RunKinds["answer-visibility"]),
-        inArray8(runs.status, [RunStatuses.completed, RunStatuses.partial]),
+        inArray9(runs.status, [RunStatuses.completed, RunStatuses.partial]),
         notProbeRun(),
         lt4(runs.createdAt, cutoff)
       );
@@ -18483,7 +18919,7 @@ async function analyticsRoutes(app) {
             provider: querySnapshots.provider,
             model: querySnapshots.model,
             servedModel: querySnapshots.servedModel
-          }).from(querySnapshots).where(inArray8(querySnapshots.runId, runIds2)).all())) {
+          }).from(querySnapshots).where(inArray9(querySnapshots.runId, runIds2)).all())) {
             if (pending.has(snapshot.provider)) {
               const models = modelsByProvider.get(snapshot.provider) ?? [];
               models.push(snapshot.model);
@@ -18629,7 +19065,7 @@ async function analyticsRoutes(app) {
         citationState: querySnapshots.citationState,
         answerMentioned: querySnapshots.answerMentioned,
         answerText: querySnapshots.answerText
-      }).from(querySnapshots).where(inArray8(querySnapshots.runId, windowRunIds)).all());
+      }).from(querySnapshots).where(inArray9(querySnapshots.runId, windowRunIds)).all());
       for (const s of allWindowSnaps) {
         const timePoint = runIdToCreatedAt.get(s.runId) ?? s.runId;
         let entry = consistencyMap.get(s.queryId);
@@ -18650,7 +19086,7 @@ async function analyticsRoutes(app) {
       answerMentioned: querySnapshots.answerMentioned,
       answerText: querySnapshots.answerText,
       competitorOverlap: querySnapshots.competitorOverlap
-    }).from(querySnapshots).leftJoin(queries, eq23(querySnapshots.queryId, queries.id)).where(inArray8(querySnapshots.runId, latestGroupRunIds)).all());
+    }).from(querySnapshots).leftJoin(queries, eq23(querySnapshots.queryId, queries.id)).where(inArray9(querySnapshots.runId, latestGroupRunIds)).all());
     const snapshots = rawSnapshots.map((s) => ({
       ...s,
       resolvedMentioned: resolveSnapshotAnswerMentioned(s, project)
@@ -18769,7 +19205,7 @@ async function analyticsRoutes(app) {
       query: queries.query,
       provider: querySnapshots.provider,
       rawResponse: querySnapshots.rawResponse
-    }).from(querySnapshots).leftJoin(queries, eq23(querySnapshots.queryId, queries.id)).where(inArray8(querySnapshots.runId, windowRunIds)).all();
+    }).from(querySnapshots).leftJoin(queries, eq23(querySnapshots.queryId, queries.id)).where(inArray9(querySnapshots.runId, windowRunIds)).all();
     const overallCounts = /* @__PURE__ */ new Map();
     const byQuery = {};
     const overallDomains = /* @__PURE__ */ new Map();
@@ -19041,7 +19477,7 @@ function buildCategoryCounts(counts2) {
 }
 
 // ../api-routes/src/intelligence.ts
-import { eq as eq24, desc as desc9, and as and18, inArray as inArray9, like } from "drizzle-orm";
+import { eq as eq24, desc as desc9, and as and18, inArray as inArray10, like } from "drizzle-orm";
 var SEVERITY_RANK = { low: 0, medium: 1, high: 2, critical: 3 };
 function severitiesAtOrAbove(min) {
   const floor = SEVERITY_RANK[min];
@@ -19171,7 +19607,7 @@ async function intelligenceRoutes(app) {
     }
     const severityFilter = request.query.severity?.trim();
     if (severityFilter) {
-      conditions.push(inArray9(insights.severity, severitiesAtOrAbove(severityFilter)));
+      conditions.push(inArray10(insights.severity, severitiesAtOrAbove(severityFilter)));
     }
     let limit;
     if (request.query.limit !== void 0) {
@@ -19209,7 +19645,7 @@ async function intelligenceRoutes(app) {
     const projectVisRuns = app.db.select({ id: runs.id, createdAt: runs.createdAt }).from(runs).where(and18(
       eq24(runs.projectId, project.id),
       eq24(runs.kind, RunKinds["answer-visibility"]),
-      inArray9(runs.status, [RunStatuses.completed, RunStatuses.partial]),
+      inArray10(runs.status, [RunStatuses.completed, RunStatuses.partial]),
       // Health-latest is the dashboard headline; probe runs must not
       // displace the most recent real visibility sweep.
       notProbeRun()
@@ -19219,7 +19655,7 @@ async function intelligenceRoutes(app) {
     if (latestGroupRunIds.length > 0) {
       const groupRows = app.db.select().from(healthSnapshots).where(and18(
         eq24(healthSnapshots.projectId, project.id),
-        inArray9(healthSnapshots.runId, latestGroupRunIds)
+        inArray10(healthSnapshots.runId, latestGroupRunIds)
       )).all();
       if (groupRows.length > 0) {
         return reply.send(aggregateHealthSnapshots(project.id, groupRows));
@@ -19241,14 +19677,14 @@ async function intelligenceRoutes(app) {
 }
 
 // ../api-routes/src/report.ts
-import { and as and22, desc as desc12, eq as eq28, gte as gte5, inArray as inArray11, lt as lt5, lte as lte3, ne as ne3, or as or4, sql as sql9 } from "drizzle-orm";
+import { and as and22, desc as desc12, eq as eq28, gte as gte5, inArray as inArray12, lt as lt5, lte as lte3, ne as ne3, or as or4, sql as sql9 } from "drizzle-orm";
 
 // ../api-routes/src/content.ts
 import crypto21 from "crypto";
 import { and as and20, desc as desc11, eq as eq26 } from "drizzle-orm";
 
 // ../api-routes/src/content-data.ts
-import { and as and19, eq as eq25, desc as desc10, inArray as inArray10 } from "drizzle-orm";
+import { and as and19, eq as eq25, desc as desc10, inArray as inArray11 } from "drizzle-orm";
 var RECENT_RUNS_WINDOW = 5;
 function loadOrchestratorInput(db, project, locationFilter = void 0) {
   const projectId = project.id;
@@ -19384,7 +19820,7 @@ function listRecentAnswerVisibilityRunIds(db, projectId, limit, locationFilter)
       // Queued/running/failed/cancelled runs may have partial or no
       // snapshots; including them risks pointing latestRunId at a run with
       // no usable evidence.
-      inArray10(runs.status, [RunStatuses.completed, RunStatuses.partial]),
+      inArray11(runs.status, [RunStatuses.completed, RunStatuses.partial]),
       // Probe runs are operator/agent test runs; they must not poison the
       // recent-runs window the content engine uses to recommend actions.
       notProbeRun()
@@ -19434,7 +19870,7 @@ function buildCandidateQueries(opts) {
   const queryRows = opts.db.select({ id: queries.id, text: queries.query }).from(queries).where(eq25(queries.projectId, opts.projectId)).all();
   const queryIdByText = new Map(queryRows.map((r) => [r.text, r.id]));
   const candidateQueryIds = opts.candidateQueryStrings.map((q) => queryIdByText.get(q)).filter((id) => Boolean(id));
-  const snapshotRows = filterTrackedSnapshots(opts.db.select().from(querySnapshots).where(inArray10(querySnapshots.runId, opts.recentRunIds)).all()).filter((r) => candidateQueryIds.includes(r.queryId));
+  const snapshotRows = filterTrackedSnapshots(opts.db.select().from(querySnapshots).where(inArray11(querySnapshots.runId, opts.recentRunIds)).all()).filter((r) => candidateQueryIds.includes(r.queryId));
   const snapshotsByQuery = /* @__PURE__ */ new Map();
   for (const row of snapshotRows) {
     const list = snapshotsByQuery.get(row.queryId) ?? [];
@@ -22361,7 +22797,7 @@ function loadSnapshotsForRun(db, runId) {
 }
 function loadSnapshotsForRunIds(db, runIds) {
   if (runIds.length === 0) return [];
-  const rows = db.select().from(querySnapshots).where(inArray11(querySnapshots.runId, [...runIds])).all();
+  const rows = db.select().from(querySnapshots).where(inArray12(querySnapshots.runId, [...runIds])).all();
   return rows.filter((r) => r.queryId !== null).map((r) => ({
     id: r.id,
     runId: r.runId,
@@ -23110,7 +23546,7 @@ function buildInsightList(db, projectId, locationFilter) {
     )
   ).orderBy(desc12(runs.createdAt)).all().filter((r) => locationFilter === void 0 || (r.location ?? null) === locationFilter).slice(0, INSIGHT_LOOKBACK_RUNS).map((r) => r.id);
   if (recentRunIds.length === 0) return [];
-  const rows = db.select().from(insights).where(and22(eq28(insights.projectId, projectId), inArray11(insights.runId, recentRunIds))).orderBy(desc12(insights.createdAt)).all();
+  const rows = db.select().from(insights).where(and22(eq28(insights.projectId, projectId), inArray12(insights.runId, recentRunIds))).orderBy(desc12(insights.createdAt)).all();
   const severityRank = { critical: 0, high: 1, medium: 2, low: 3 };
   const flat = rows.filter((r) => !r.dismissed).map((r) => {
     const recommendation = r.recommendation;
@@ -24859,7 +25295,7 @@ async function organicEvidenceRoutes(app) {
 }
 
 // ../api-routes/src/citations.ts
-import { and as and25, eq as eq31, inArray as inArray12 } from "drizzle-orm";
+import { and as and25, eq as eq31, inArray as inArray13 } from "drizzle-orm";
 async function citationRoutes(app) {
   app.get("/projects/:name/citations/visibility", async (request, reply) => {
     const project = resolveProject(app.db, request.params.name);
@@ -24883,7 +25319,7 @@ async function citationRoutes(app) {
       competitorOverlap: querySnapshots.competitorOverlap,
       answerMentioned: querySnapshots.answerMentioned,
       createdAt: querySnapshots.createdAt
-    }).from(querySnapshots).where(inArray12(querySnapshots.runId, projectRuns.map((r) => r.id))).all();
+    }).from(querySnapshots).where(inArray13(querySnapshots.runId, projectRuns.map((r) => r.id))).all();
     if (rawSnapshots.length === 0) {
       return reply.send(emptyCitationVisibility("no-runs-yet"));
     }
@@ -25022,7 +25458,7 @@ function computeCitationVisibility(input) {
 }
 
 // ../api-routes/src/visibility-stats.ts
-import { and as and26, desc as desc14, eq as eq32, inArray as inArray13 } from "drizzle-orm";
+import { and as and26, desc as desc14, eq as eq32, inArray as inArray14 } from "drizzle-orm";
 
 // ../api-routes/src/visibility-compare.ts
 var VISIBILITY_COMPARE_MIN_RUNS = 5;
@@ -25433,7 +25869,7 @@ async function visibilityStatsRoutes(app) {
       citationState: querySnapshots.citationState,
       answerMentioned: querySnapshots.answerMentioned,
       createdAt: querySnapshots.createdAt
-    }).from(querySnapshots).where(inArray13(querySnapshots.runId, runIds)).all() : [];
+    }).from(querySnapshots).where(inArray14(querySnapshots.runId, runIds)).all() : [];
     const stats = computeVisibilityStats({ queries: projectQueries, snapshots, groupBy });
     const queryAttribution = buildQueryAttribution(projectQueries);
     let shareOfVoice;
@@ -25448,7 +25884,7 @@ async function visibilityStatsRoutes(app) {
         queryText: querySnapshots.queryText,
         answerMentioned: querySnapshots.answerMentioned,
         answerText: querySnapshots.answerText
-      }).from(querySnapshots).where(inArray13(querySnapshots.runId, runIds)).all() : [];
+      }).from(querySnapshots).where(inArray14(querySnapshots.runId, runIds)).all() : [];
       const attributedSovSnapshots = sovSnapshots.filter((s) => resolveCurrentQuery(queryAttribution, s) !== void 0);
       const result = buildMentionShare(
         attributedSovSnapshots.map((s) => ({ projectMentioned: s.answerMentioned === true, answerText: s.answerText })),
@@ -25529,7 +25965,7 @@ async function visibilityStatsRoutes(app) {
         answerMentioned: querySnapshots.answerMentioned,
         answerText: querySnapshots.answerText,
         citedDomains: querySnapshots.citedDomains
-      }).from(querySnapshots).where(inArray13(querySnapshots.runId, runIds)).all() : [];
+      }).from(querySnapshots).where(inArray14(querySnapshots.runId, runIds)).all() : [];
       return { runCount: monthRuns.length, snapshots };
     };
     const fromMonth = loadMonth(fromBounds);
@@ -25784,7 +26220,7 @@ async function resultsExportRoutes(app) {
 }
 
 // ../api-routes/src/composites.ts
-import { eq as eq34, and as and28, desc as desc15, sql as sql12, like as like2, or as or7, inArray as inArray14 } from "drizzle-orm";
+import { eq as eq34, and as and28, desc as desc15, sql as sql12, like as like2, or as or7, inArray as inArray15 } from "drizzle-orm";
 var TOP_INSIGHT_LIMIT = 5;
 var SEARCH_HIT_HARD_LIMIT = 50;
 var SEARCH_SNIPPET_RADIUS = 80;
@@ -26051,7 +26487,7 @@ function loadSnapshotsByRunIds(app, runIds, queryIdByText) {
     answerText: querySnapshots.answerText,
     competitorOverlap: querySnapshots.competitorOverlap,
     citedDomains: querySnapshots.citedDomains
-  }).from(querySnapshots).where(inArray14(querySnapshots.runId, [...runIds])).all();
+  }).from(querySnapshots).where(inArray15(querySnapshots.runId, [...runIds])).all();
   for (const row of rows) {
     const queryText = row.queryText?.trim() || null;
     let queryId;
@@ -26686,6 +27122,12 @@ var SCHEMA_TABLE = {
   SiteAuditRunResponseDto: siteAuditRunResponseSchema,
   SiteAuditScoreDto: siteAuditScoreSchema,
   SiteAuditTrendResponseDto: siteAuditTrendResponseSchema,
+  SiteCrawlDeadLinksResponseDto: siteCrawlDeadLinksResponseSchema,
+  SiteCrawlInternalLinksResponseDto: siteCrawlInternalLinksResponseSchema,
+  SiteCrawlNeighborsResponseDto: siteCrawlNeighborsResponseSchema,
+  SiteCrawlPagesResponseDto: siteCrawlPagesResponseSchema,
+  SiteCrawlStructureResponseDto: siteCrawlStructureResponseSchema,
+  SiteCrawlSummaryDto: siteCrawlSummarySchema,
   SnapshotDiffResponse: snapshotDiffResponseSchema,
   SnapshotListResponse: snapshotListResponseSchema,
   SnapshotReportDto: snapshotReportSchema,
@@ -27142,6 +27584,18 @@ var measurementCursorParameter = {
   description: "Opaque cursor from the previous page.",
   schema: stringSchema
 };
+var crawlCursorParameter = {
+  name: "cursor",
+  in: "query",
+  description: "Opaque cursor from the previous bounded crawl response.",
+  schema: stringSchema
+};
+var crawlLimitParameter = {
+  name: "limit",
+  in: "query",
+  description: "Page size. Defaults to 100 (50 for structure/neighbors); maximum 200 (100 for structure/neighbors).",
+  schema: { type: "integer", minimum: 1, maximum: 200 }
+};
 var measurementOverviewSortParameter = {
   name: "sort",
   in: "query",
@@ -32593,11 +33047,122 @@ var routeCatalog = [
       404: errorResponse("Project not found.")
     }
   },
+  {
+    method: "get",
+    path: "/api/v1/projects/{name}/technical-aeo/crawl",
+    summary: "Get persisted Technical AEO crawl metadata",
+    description: "Returns the latest complete non-probe site-audit crawl, or the selected historical run (which may be partial). `hasCrawlData=false` never synthesizes a graph from legacy scorecard rows; `legacyAuditAvailable` says that the old score/pages/trend data can still be read separately.",
+    tags: ["technical-aeo"],
+    parameters: [
+      nameParameter,
+      { name: "runId", in: "query", description: "Historical site-audit run ID. Omit for the latest crawl.", schema: stringSchema }
+    ],
+    responses: {
+      200: jsonResponse("Persisted crawl summary returned.", "SiteCrawlSummaryDto"),
+      404: errorResponse("Project or crawl-bearing site-audit run not found.")
+    }
+  },
+  {
+    method: "get",
+    path: "/api/v1/projects/{name}/technical-aeo/crawl/pages",
+    summary: "List persisted Technical AEO crawl pages",
+    description: "Cursor-paged canonical crawl nodes for the latest or selected crawl. `inventoryEligible` is Canonry technical-inventory eligibility, not a statement about actual Google index state.",
+    tags: ["technical-aeo"],
+    parameters: [
+      nameParameter,
+      { name: "runId", in: "query", description: "Historical site-audit run ID. Omit for the latest crawl.", schema: stringSchema },
+      { name: "inventoryEligible", in: "query", description: "Filter Canonry technical inventory eligibility (`true` or `false`).", schema: booleanSchema },
+      { name: "fetchState", in: "query", description: "Filter by persisted fetch state.", schema: stringSchema },
+      { name: "indexabilityState", in: "query", description: "Filter by crawler-derived indexability state; this is not Google index coverage.", schema: stringSchema },
+      { name: "auditState", in: "query", description: "Filter by audit state.", schema: stringSchema },
+      { name: "sort", in: "query", description: "Sort order. Defaults to `url`.", schema: { type: "string", enum: ["url", "path", "score-asc", "score-desc"] } },
+      crawlCursorParameter,
+      crawlLimitParameter
+    ],
+    responses: {
+      200: jsonResponse("Crawl pages returned.", "SiteCrawlPagesResponseDto"),
+      404: errorResponse("Project or crawl-bearing site-audit run not found.")
+    }
+  },
+  {
+    method: "get",
+    path: "/api/v1/projects/{name}/technical-aeo/structure",
+    summary: "List one level of Technical AEO crawl structure",
+    description: "Returns immediate children under `parentPath` only. It is deliberately bounded and never emits a full site tree.",
+    tags: ["technical-aeo"],
+    parameters: [
+      nameParameter,
+      { name: "runId", in: "query", description: "Historical site-audit run ID. Omit for the latest crawl.", schema: stringSchema },
+      { name: "parentPath", in: "query", description: "Parent URL path. Defaults to `/`.", schema: stringSchema },
+      crawlCursorParameter,
+      { name: "limit", in: "query", description: "Maximum immediate children. Defaults to 50; maximum 100.", schema: { type: "integer", minimum: 1, maximum: 100 } }
+    ],
+    responses: {
+      200: jsonResponse("Immediate crawl children returned.", "SiteCrawlStructureResponseDto"),
+      404: errorResponse("Project or crawl-bearing site-audit run not found.")
+    }
+  },
+  {
+    method: "get",
+    path: "/api/v1/projects/{name}/technical-aeo/internal-links",
+    summary: "List persisted internal crawl links",
+    description: "Cursor-paged internal edges for the latest or selected crawl. Optional source/target/followability filters remain project-, run-, and attempt-scoped.",
+    tags: ["technical-aeo"],
+    parameters: [
+      nameParameter,
+      { name: "runId", in: "query", description: "Historical site-audit run ID. Omit for the latest crawl.", schema: stringSchema },
+      { name: "sourceUrl", in: "query", description: "Restrict to a source URL.", schema: stringSchema },
+      { name: "targetUrl", in: "query", description: "Restrict to a target URL.", schema: stringSchema },
+      { name: "followable", in: "query", description: "Restrict to followable or nofollow link observations.", schema: booleanSchema },
+      crawlCursorParameter,
+      crawlLimitParameter
+    ],
+    responses: {
+      200: jsonResponse("Internal crawl links returned.", "SiteCrawlInternalLinksResponseDto"),
+      404: errorResponse("Project or crawl-bearing site-audit run not found.")
+    }
+  },
+  {
+    method: "get",
+    path: "/api/v1/projects/{name}/technical-aeo/internal-links/neighbors",
+    summary: "Get a bounded internal-link neighborhood",
+    description: "Returns bounded inbound and outbound internal edges for one `nodeKey` or URL; this is not a full graph traversal.",
+    tags: ["technical-aeo"],
+    parameters: [
+      nameParameter,
+      { name: "runId", in: "query", description: "Historical site-audit run ID. Omit for the latest crawl.", schema: stringSchema },
+      { name: "nodeKey", in: "query", description: "Canonical crawl node key.", schema: stringSchema },
+      { name: "url", in: "query", description: "Canonical crawl URL.", schema: stringSchema },
+      { name: "limit", in: "query", description: "Maximum inbound and outbound edges independently. Defaults to 50; maximum 100.", schema: { type: "integer", minimum: 1, maximum: 100 } }
+    ],
+    responses: {
+      200: jsonResponse("Bounded internal-link neighborhood returned.", "SiteCrawlNeighborsResponseDto"),
+      400: errorResponse("A nodeKey or URL is required."),
+      404: errorResponse("Project or crawl-bearing site-audit run not found.")
+    }
+  },
+  {
+    method: "get",
+    path: "/api/v1/projects/{name}/technical-aeo/dead-links",
+    summary: "List dead-link findings when checks were enabled",
+    description: "Returns a discriminated `disabled`, `complete`, `partial`, or `unavailable` state. A disabled check is never returned as an empty zero-result list.",
+    tags: ["technical-aeo"],
+    parameters: [
+      nameParameter,
+      { name: "runId", in: "query", description: "Historical site-audit run ID. Omit for the latest crawl.", schema: stringSchema },
+      crawlCursorParameter,
+      crawlLimitParameter
+    ],
+    responses: {
+      200: jsonResponse("Dead-link status and findings returned.", "SiteCrawlDeadLinksResponseDto"),
+      404: errorResponse("Project or crawl-bearing site-audit run not found.")
+    }
+  },
   {
     method: "post",
     path: "/api/v1/projects/{name}/technical-aeo/runs",
     summary: "Trigger a Technical AEO site-audit run",
-    description: "Queues a `site-audit` run that crawls the project sitemap and audits every reachable page. Idempotent: if a site-audit run is already queued/running for the project, returns that run instead of starting a second.",
+    description: "Queues a `site-audit` run. The run discovers pages from the root, recursive sitemaps, and internal links. An active run is reused only when its persisted effective sitemap, budgets, depth, and dead-link option match exactly; a semantic mismatch returns 409.",
     tags: ["technical-aeo"],
     parameters: [nameParameter],
     requestBody: {
@@ -32607,8 +33172,12 @@ var routeCatalog = [
           schema: {
             type: "object",
             properties: {
-              sitemapUrl: { ...stringSchema, description: "Override the sitemap URL. Defaults to https://<canonicalDomain>/sitemap.xml." },
-              limit: { ...integerSchema, description: "Cap pages audited (highest sitemap <priority> first). Max 2000." }
+              sitemapUrl: { ...stringSchema, description: "Deprecated compatibility alias for the sitemap override." },
+              limit: { ...integerSchema, minimum: 1, maximum: 2e3, description: "Deprecated compatibility alias for maxPages. Max 2000." },
+              maxPages: { ...integerSchema, minimum: 1, maximum: 5e4, description: "Crawl page budget. Defaults to 1000; max 50000." },
+              maxEdges: { ...integerSchema, minimum: 1, maximum: 1e6, description: "Internal-link observation budget. Defaults to 100000; max 1000000." },
+              maxDepth: { ...integerSchema, minimum: 0, maximum: 100, description: "Maximum crawl depth from the root. Max 100." },
+              checkDeadLinks: { ...booleanSchema, default: false, description: "Enable dead-link checking. Defaults to false." }
             }
           }
         }
@@ -32617,6 +33186,7 @@ var routeCatalog = [
     responses: {
       200: jsonResponse("Site-audit run queued (or the in-flight run returned).", "SiteAuditRunResponseDto"),
       400: errorResponse("Invalid site-audit request."),
+      409: errorResponse("A crawl with different effective options is already queued or running."),
       404: errorResponse("Project not found.")
     }
   }
@@ -33654,7 +34224,7 @@ function formatNotification(row) {
 
 // ../api-routes/src/google.ts
 import crypto29 from "crypto";
-import { eq as eq39, and as and30, desc as desc17, sql as sql13, inArray as inArray15 } from "drizzle-orm";
+import { eq as eq39, and as and30, desc as desc17, sql as sql13, inArray as inArray16 } from "drizzle-orm";
 
 // ../api-routes/src/gbp-summary.ts
 function computeMetricTotals(rows) {
@@ -36914,10 +37484,10 @@ async function googleRoutes(app, opts) {
         locationProfiles: []
       });
     }
-    const metricRows = app.db.select().from(gbpDailyMetrics).where(and30(eq39(gbpDailyMetrics.projectId, project.id), inArray15(gbpDailyMetrics.locationName, locationNames))).all();
-    const keywordRows = app.db.select().from(gbpKeywordImpressions).where(and30(eq39(gbpKeywordImpressions.projectId, project.id), inArray15(gbpKeywordImpressions.locationName, locationNames))).all();
-    const placeActionRows = app.db.select().from(gbpPlaceActions).where(and30(eq39(gbpPlaceActions.projectId, project.id), inArray15(gbpPlaceActions.locationName, locationNames))).all();
-    const lodgingRows = app.db.select().from(gbpLodgingSnapshots).where(and30(eq39(gbpLodgingSnapshots.projectId, project.id), inArray15(gbpLodgingSnapshots.locationName, locationNames))).orderBy(desc17(gbpLodgingSnapshots.syncedAt)).all();
+    const metricRows = app.db.select().from(gbpDailyMetrics).where(and30(eq39(gbpDailyMetrics.projectId, project.id), inArray16(gbpDailyMetrics.locationName, locationNames))).all();
+    const keywordRows = app.db.select().from(gbpKeywordImpressions).where(and30(eq39(gbpKeywordImpressions.projectId, project.id), inArray16(gbpKeywordImpressions.locationName, locationNames))).all();
+    const placeActionRows = app.db.select().from(gbpPlaceActions).where(and30(eq39(gbpPlaceActions.projectId, project.id), inArray16(gbpPlaceActions.locationName, locationNames))).all();
+    const lodgingRows = app.db.select().from(gbpLodgingSnapshots).where(and30(eq39(gbpLodgingSnapshots.projectId, project.id), inArray16(gbpLodgingSnapshots.locationName, locationNames))).orderBy(desc17(gbpLodgingSnapshots.syncedAt)).all();
     const latestLodgingByLocation = /* @__PURE__ */ new Map();
     for (const row of lodgingRows) {
       if (!latestLodgingByLocation.has(row.locationName)) {
@@ -36931,7 +37501,7 @@ async function googleRoutes(app, opts) {
       regularHours: gbpLocations.regularHours,
       primaryPhone: gbpLocations.primaryPhone,
       openStatus: gbpLocations.openStatus
-    }).from(gbpLocations).where(and30(eq39(gbpLocations.projectId, project.id), inArray15(gbpLocations.locationName, locationNames))).all();
+    }).from(gbpLocations).where(and30(eq39(gbpLocations.projectId, project.id), inArray16(gbpLocations.locationName, locationNames))).all();
     return buildGbpSummary({
       locationName,
       locationCount: locationNames.length,
@@ -36947,7 +37517,7 @@ async function googleRoutes(app, opts) {
 
 // ../api-routes/src/ads.ts
 import crypto32 from "crypto";
-import { eq as eq41, and as and32, asc as asc8, desc as desc18, gt as gt2, gte as gte9, lt as lt6, lte as lte8, ne as ne4, inArray as inArray16, or as or9, isNull as isNull3, isNotNull as isNotNull2, sql as sql15 } from "drizzle-orm";
+import { eq as eq41, and as and32, asc as asc8, desc as desc18, gt as gt2, gte as gte9, lt as lt6, lte as lte8, ne as ne4, inArray as inArray17, or as or9, isNull as isNull3, isNotNull as isNotNull2, sql as sql15 } from "drizzle-orm";
 
 // ../api-routes/src/ads-activation-routes.ts
 import crypto31 from "crypto";
@@ -39894,7 +40464,7 @@ function semanticallyMatchesMaterializedEntity(app, input) {
     eq41(adsOperations.projectId, input.projectId),
     eq41(adsOperations.entityType, input.entityType),
     eq41(adsOperations.entityId, input.entityId),
-    inArray16(adsOperations.kind, [createKind, updateKind])
+    inArray17(adsOperations.kind, [createKind, updateKind])
   )).orderBy(asc8(adsOperations.createdAt), asc8(adsOperations.id)).all();
   if (rows.some((row) => row.state === AdsOperationStates.pending || row.state === AdsOperationStates.reconciling || row.state === AdsOperationStates.unknown || row.kind === updateKind && row.state === AdsOperationStates.succeeded)) {
     return false;
@@ -40331,7 +40901,7 @@ function quarantineExhaustedReconciliation(app, row, now, policy, context) {
       eq41(adsOperations.id, row.id),
       eq41(adsOperations.projectId, row.projectId),
       gte9(adsOperations.reconcileAttempts, policy.maxAttempts),
-      inArray16(adsOperations.state, [
+      inArray17(adsOperations.state, [
         AdsOperationStates.pending,
         AdsOperationStates.unknown,
         AdsOperationStates.reconciling
@@ -40555,7 +41125,7 @@ async function sweepAdsOperationsOnce(app, opts, config) {
         eq41(adsOperations.reconcileStrategy, AdsReconcileStrategies.create_fingerprint),
         or9(
           isNotNull2(adsOperations.entityId),
-          inArray16(adsOperations.state, [
+          inArray17(adsOperations.state, [
             AdsOperationStates.pending,
             AdsOperationStates.reconciling
           ])
@@ -40603,7 +41173,7 @@ async function sweepAdsOperationsOnce(app, opts, config) {
   const eligibleProjectIds = [...credentialsByProjectId.keys()];
   if (eligibleProjectIds.length === 0) return;
   const rows = app.db.select().from(adsOperations).where(and32(
-    inArray16(adsOperations.projectId, eligibleProjectIds),
+    inArray17(adsOperations.projectId, eligibleProjectIds),
     staleReconcileWhere
   )).orderBy(asc8(adsOperations.updatedAt)).limit(config.batchSize).all();
   for (const row of rows) {
@@ -40931,7 +41501,7 @@ async function adsRoutes(app, opts) {
     const cursor = parsed.data.cursor ? decodeAdsOperationCursor(parsed.data.cursor, project.id, parsed.data.state) : void 0;
     const rows = app.db.select().from(adsOperations).where(and32(
       eq41(adsOperations.projectId, project.id),
-      inArray16(adsOperations.state, parsed.data.state),
+      inArray17(adsOperations.state, parsed.data.state),
       cursor ? or9(
         gt2(adsOperations.createdAt, cursor.createdAt),
         and32(
@@ -41278,7 +41848,7 @@ async function adsRoutes(app, opts) {
     const inFlight = app.db.select({ id: runs.id, status: runs.status }).from(runs).where(and32(
       eq41(runs.projectId, project.id),
       eq41(runs.kind, RunKinds["ads-sync"]),
-      inArray16(runs.status, [RunStatuses.queued, RunStatuses.running])
+      inArray17(runs.status, [RunStatuses.queued, RunStatuses.running])
     )).get();
     if (inFlight) {
       const existing = { runId: inFlight.id, status: inFlight.status };
@@ -54135,7 +54705,7 @@ async function doctorRoutes(app, opts) {
 
 // ../api-routes/src/discovery/routes.ts
 import crypto41 from "crypto";
-import { and as and42, desc as desc23, eq as eq53, gte as gte12, inArray as inArray17, isNull as isNull4, or as or10 } from "drizzle-orm";
+import { and as and42, desc as desc23, eq as eq53, gte as gte12, inArray as inArray18, isNull as isNull4, or as or10 } from "drizzle-orm";
 var MAX_INFLIGHT_DISCOVERY_AGE_MS = 2 * 60 * 60 * 1e3;
 async function discoveryRoutes(app, opts) {
   app.post("/projects/:name/discover/run", async (request, reply) => {
@@ -54190,7 +54760,7 @@ async function discoveryRoutes(app, opts) {
         // must never reuse another set's session. Null (= Gemini-only
         // default, including explicit ['gemini']) matches legacy rows.
         seedProviders == null ? isNull4(discoverySessions.seedProviders) : eq53(discoverySessions.seedProviders, seedProviders),
-        inArray17(discoverySessions.status, [
+        inArray18(discoverySessions.status, [
           DiscoverySessionStatuses.queued,
           DiscoverySessionStatuses.seeding,
           DiscoverySessionStatuses.probing
@@ -54815,7 +55385,7 @@ function dedupeStrings(input) {
 
 // ../api-routes/src/technical-aeo.ts
 import crypto43 from "crypto";
-import { and as and43, asc as asc10, count, desc as desc24, eq as eq55, inArray as inArray18, lt as lt7 } from "drizzle-orm";
+import { and as and43, asc as asc10, count, desc as desc24, eq as eq55, inArray as inArray19, lt as lt7, or as or11 } from "drizzle-orm";
 var SURFACEABLE_STATUSES = [RunStatuses.completed, RunStatuses.partial];
 function emptyScore(projectName) {
   return {
@@ -54844,13 +55414,133 @@ function parsePositiveInt(value, fallback, max) {
   if (!Number.isFinite(n) || n < 0) return fallback;
   return Math.min(max, Math.floor(n));
 }
+function parseBoundedLimit(value, fallback, max) {
+  const parsed = parsePositiveInt(value, fallback, max);
+  return Math.max(1, parsed);
+}
+function parseBoolean2(value) {
+  if (value === true || value === "true" || value === "1") return true;
+  if (value === false || value === "false" || value === "0") return false;
+  return null;
+}
+function encodeCursor2(offset) {
+  return Buffer.from(JSON.stringify({ offset })).toString("base64url");
+}
+function decodeCursor(value) {
+  if (typeof value !== "string" || value.length === 0 || value.length > 256) return 0;
+  try {
+    const parsed = JSON.parse(Buffer.from(value, "base64url").toString("utf8"));
+    return typeof parsed.offset === "number" && Number.isSafeInteger(parsed.offset) && parsed.offset >= 0 ? Math.min(parsed.offset, 1e6) : 0;
+  } catch {
+    return 0;
+  }
+}
+function normalizeParentPath(value) {
+  if (typeof value !== "string" || value.trim() === "") return "/";
+  const bare = value.trim().split(/[?#]/, 1)[0] ?? "/";
+  const prefixed = bare.startsWith("/") ? bare : `/${bare}`;
+  const collapsed = prefixed.replace(/\/{2,}/g, "/");
+  return collapsed.length > 1 && collapsed.endsWith("/") ? collapsed.slice(0, -1) : collapsed;
+}
+var MAX_STRUCTURE_SOURCE_ROWS = 5e4;
+function structureHierarchyPath(path8) {
+  if (path8.length <= 1) return path8;
+  return path8.replace(/\/+$/, "") || "/";
+}
+function structureChildPath(parentPath, path8) {
+  const hierarchyPath = structureHierarchyPath(path8);
+  if (hierarchyPath === parentPath) return null;
+  const relative = parentPath === "/" ? hierarchyPath.startsWith("/") ? hierarchyPath.slice(1) : "" : hierarchyPath.startsWith(`${parentPath}/`) ? hierarchyPath.slice(parentPath.length + 1) : "";
+  if (relative.length === 0) return null;
+  const separator = relative.indexOf("/");
+  const segment = separator === -1 ? relative : relative.slice(0, separator);
+  return parentPath === "/" ? `/${segment}` : `${parentPath}/${segment}`;
+}
+function mapCrawlPage(row) {
+  return {
+    nodeKey: row.nodeKey,
+    url: row.url,
+    finalUrl: row.finalUrl,
+    path: row.path,
+    parentPath: row.parentPath,
+    discoverySource: row.discoverySource,
+    fetchState: row.fetchState,
+    httpStatus: row.httpStatus,
+    canonicalUrl: row.canonicalUrl,
+    indexabilityState: row.indexabilityState,
+    indexabilityReasons: row.indexabilityReasons,
+    auditState: row.auditState,
+    auditScore: row.auditScore,
+    inventoryEligible: row.inventoryEligible,
+    depth: row.depth,
+    inboundUniqueEdges: row.inboundUniqueEdges,
+    outboundUniqueEdges: row.outboundUniqueEdges,
+    inboundOccurrences: row.inboundOccurrences,
+    outboundOccurrences: row.outboundOccurrences,
+    linkScoreRaw: row.linkScoreRaw,
+    linkScoreNormalized: row.linkScoreNormalized
+  };
+}
+function mapCrawlEdge(row) {
+  return {
+    edgeKey: row.edgeKey,
+    sourceNodeKey: row.sourceNodeKey,
+    sourceUrl: row.sourceUrl,
+    targetNodeKey: row.targetNodeKey,
+    targetUrl: row.targetUrl,
+    relation: row.relation,
+    internal: row.internal,
+    followable: row.followable,
+    occurrences: row.occurrences,
+    followableOccurrences: row.followableOccurrences,
+    nofollowOccurrences: row.nofollowOccurrences,
+    anchors: row.anchors
+  };
+}
 async function technicalAeoRoutes(app, opts) {
+  const resolveCrawl = (projectId, runId) => {
+    const filters = [
+      eq55(siteCrawlSnapshots.projectId, projectId),
+      eq55(runs.projectId, projectId),
+      eq55(runs.kind, RunKinds["site-audit"]),
+      inArray19(runs.status, SURFACEABLE_STATUSES),
+      notProbeRun()
+    ];
+    if (runId) {
+      filters.push(eq55(siteCrawlSnapshots.runId, runId));
+    } else {
+      filters.push(eq55(siteCrawlSnapshots.complete, true), eq55(runs.status, RunStatuses.completed));
+    }
+    return app.db.select({ snapshot: siteCrawlSnapshots, runStatus: runs.status }).from(siteCrawlSnapshots).innerJoin(runs, eq55(siteCrawlSnapshots.runId, runs.id)).where(and43(...filters)).orderBy(desc24(siteCrawlSnapshots.createdAt)).limit(1).get();
+  };
+  const hasLegacyAudit = (projectId) => Boolean(app.db.select({ runId: siteAuditSnapshots.runId }).from(siteAuditSnapshots).innerJoin(runs, eq55(siteAuditSnapshots.runId, runs.id)).where(and43(
+    eq55(siteAuditSnapshots.projectId, projectId),
+    eq55(runs.projectId, projectId),
+    eq55(runs.kind, RunKinds["site-audit"]),
+    inArray19(runs.status, SURFACEABLE_STATUSES),
+    notProbeRun()
+  )).limit(1).get());
+  const emptyCrawlSummary = (projectName, legacyAuditAvailable) => ({
+    project: projectName,
+    hasCrawlData: false,
+    legacyAuditAvailable,
+    runId: null,
+    runStatus: null,
+    rootUrl: null,
+    effectiveOptions: {},
+    complete: false,
+    termination: null,
+    detailsAvailable: false,
+    counts: { pagesDiscovered: 0, pagesFetched: 0, pagesEligible: 0, edges: 0, findings: 0 },
+    deadLinks: { state: "unavailable" }
+  });
   app.get("/projects/:name/technical-aeo", async (request) => {
     const project = resolveProject(app.db, request.params.name);
     const baseFilters = [
       eq55(siteAuditSnapshots.projectId, project.id),
+      eq55(runs.projectId, project.id),
       eq55(runs.kind, RunKinds["site-audit"]),
-      inArray18(runs.status, SURFACEABLE_STATUSES),
+      inArray19(runs.status, SURFACEABLE_STATUSES),
       notProbeRun()
     ];
     const targetFilters = request.query.runId ? [...baseFilters, eq55(siteAuditSnapshots.runId, request.query.runId)] : baseFilters;
@@ -54888,8 +55578,9 @@ async function technicalAeoRoutes(app, opts) {
     const project = resolveProject(app.db, request.params.name);
     const targetFilters = [
       eq55(siteAuditSnapshots.projectId, project.id),
+      eq55(runs.projectId, project.id),
       eq55(runs.kind, RunKinds["site-audit"]),
-      inArray18(runs.status, SURFACEABLE_STATUSES),
+      inArray19(runs.status, SURFACEABLE_STATUSES),
       notProbeRun()
     ];
     if (request.query.runId) targetFilters.push(eq55(siteAuditSnapshots.runId, request.query.runId));
@@ -54901,7 +55592,7 @@ async function technicalAeoRoutes(app, opts) {
       return { project: project.name, runId: null, auditedAt: null, total: 0, pages: [] };
     }
     const statusFilter = request.query.status === "success" || request.query.status === "error" ? request.query.status : null;
-    const conds = [eq55(siteAuditPages.runId, latest.runId)];
+    const conds = [eq55(siteAuditPages.projectId, project.id), eq55(siteAuditPages.runId, latest.runId)];
     if (statusFilter) conds.push(eq55(siteAuditPages.status, statusFilter));
     const where = and43(...conds);
     const totalRow = app.db.select({ value: count() }).from(siteAuditPages).where(where).get();
@@ -54929,41 +55620,363 @@ async function technicalAeoRoutes(app, opts) {
       pagesAudited: siteAuditSnapshots.pagesAudited
     }).from(siteAuditSnapshots).innerJoin(runs, eq55(siteAuditSnapshots.runId, runs.id)).where(and43(
       eq55(siteAuditSnapshots.projectId, project.id),
+      eq55(runs.projectId, project.id),
       eq55(runs.kind, RunKinds["site-audit"]),
-      inArray18(runs.status, SURFACEABLE_STATUSES),
+      inArray19(runs.status, SURFACEABLE_STATUSES),
       notProbeRun()
     )).orderBy(desc24(siteAuditSnapshots.createdAt)).limit(limit).all();
     return { project: project.name, points: rows.reverse() };
   });
+  app.get("/projects/:name/technical-aeo/crawl", async (request) => {
+    const project = resolveProject(app.db, request.params.name);
+    const target = resolveCrawl(project.id, request.query.runId);
+    const legacyAuditAvailable = hasLegacyAudit(project.id);
+    if (!target) {
+      if (request.query.runId) throw notFound("Site crawl run", request.query.runId);
+      return emptyCrawlSummary(project.name, legacyAuditAvailable);
+    }
+    const snapshot = target.snapshot;
+    const deadLinks = !snapshot.checkDeadLinks || snapshot.deadLinkState === "disabled" ? { state: "disabled" } : snapshot.deadLinkState === "complete" ? { state: "complete", checked: snapshot.deadLinksChecked, found: snapshot.deadLinksFound } : snapshot.deadLinkState === "partial" ? { state: "partial", checked: snapshot.deadLinksChecked, found: snapshot.deadLinksFound } : { state: "unavailable" };
+    return {
+      project: project.name,
+      hasCrawlData: true,
+      legacyAuditAvailable,
+      runId: snapshot.runId,
+      runStatus: target.runStatus,
+      rootUrl: snapshot.rootUrl,
+      crawlSchemaVersion: snapshot.crawlSchemaVersion,
+      engineVersion: snapshot.engineVersion,
+      normalizationVersion: snapshot.normalizationVersion,
+      indexabilityVersion: snapshot.indexabilityVersion,
+      linkScoreVersion: snapshot.linkScoreVersion,
+      effectiveOptions: snapshot.effectiveOptions,
+      complete: snapshot.complete,
+      termination: snapshot.termination,
+      detailsAvailable: snapshot.detailsAvailable,
+      counts: {
+        pagesDiscovered: snapshot.pagesDiscovered,
+        pagesFetched: snapshot.pagesFetched,
+        pagesEligible: snapshot.pagesEligible,
+        edges: snapshot.edgesDiscovered,
+        findings: snapshot.findingsCount
+      },
+      deadLinks
+    };
+  });
+  app.get("/projects/:name/technical-aeo/crawl/pages", async (request) => {
+    const project = resolveProject(app.db, request.params.name);
+    const target = resolveCrawl(project.id, request.query.runId);
+    if (!target) {
+      if (request.query.runId) throw notFound("Site crawl run", request.query.runId);
+      return { project: project.name, hasCrawlData: false, runId: null, total: 0, nextCursor: null, pages: [] };
+    }
+    const snapshot = target.snapshot;
+    if (!snapshot.detailsAvailable || !snapshot.attemptId) {
+      return { project: project.name, hasCrawlData: true, runId: snapshot.runId, total: 0, nextCursor: null, pages: [] };
+    }
+    const filters = [
+      eq55(siteCrawlPages.projectId, project.id),
+      eq55(siteCrawlPages.runId, snapshot.runId),
+      eq55(siteCrawlPages.attemptId, snapshot.attemptId)
+    ];
+    const inventoryEligible = parseBoolean2(request.query.inventoryEligible);
+    if (inventoryEligible != null) filters.push(eq55(siteCrawlPages.inventoryEligible, inventoryEligible));
+    if (request.query.fetchState) filters.push(eq55(siteCrawlPages.fetchState, request.query.fetchState));
+    if (request.query.indexabilityState) filters.push(eq55(siteCrawlPages.indexabilityState, request.query.indexabilityState));
+    if (request.query.auditState) filters.push(eq55(siteCrawlPages.auditState, request.query.auditState));
+    const where = and43(...filters);
+    const total = app.db.select({ value: count() }).from(siteCrawlPages).where(where).get()?.value ?? 0;
+    const limit = parseBoundedLimit(request.query.limit, 100, 200);
+    const offset = decodeCursor(request.query.cursor);
+    const orderBy = request.query.sort === "score-desc" ? [desc24(siteCrawlPages.auditScore), asc10(siteCrawlPages.nodeKey)] : request.query.sort === "score-asc" ? [asc10(siteCrawlPages.auditScore), asc10(siteCrawlPages.nodeKey)] : request.query.sort === "path" ? [asc10(siteCrawlPages.path), asc10(siteCrawlPages.nodeKey)] : [asc10(siteCrawlPages.url), asc10(siteCrawlPages.nodeKey)];
+    const rows = app.db.select().from(siteCrawlPages).where(where).orderBy(...orderBy).limit(limit).offset(offset).all();
+    const nextOffset = offset + rows.length;
+    return {
+      project: project.name,
+      hasCrawlData: true,
+      runId: snapshot.runId,
+      total,
+      nextCursor: nextOffset < total ? encodeCursor2(nextOffset) : null,
+      pages: rows.map(mapCrawlPage)
+    };
+  });
+  app.get("/projects/:name/technical-aeo/structure", async (request) => {
+    const project = resolveProject(app.db, request.params.name);
+    const parentPath = normalizeParentPath(request.query.parentPath);
+    const target = resolveCrawl(project.id, request.query.runId);
+    if (!target) {
+      if (request.query.runId) throw notFound("Site crawl run", request.query.runId);
+      return { project: project.name, hasCrawlData: false, runId: null, parentPath, nextCursor: null, children: [] };
+    }
+    const snapshot = target.snapshot;
+    if (!snapshot.detailsAvailable || !snapshot.attemptId) {
+      return { project: project.name, hasCrawlData: true, runId: snapshot.runId, parentPath, nextCursor: null, children: [] };
+    }
+    const sourceRows = app.db.select({
+      path: siteCrawlPages.path,
+      url: siteCrawlPages.url,
+      inventoryEligible: siteCrawlPages.inventoryEligible,
+      fetchState: siteCrawlPages.fetchState
+    }).from(siteCrawlPages).where(and43(
+      eq55(siteCrawlPages.projectId, project.id),
+      eq55(siteCrawlPages.runId, snapshot.runId),
+      eq55(siteCrawlPages.attemptId, snapshot.attemptId)
+    )).limit(MAX_STRUCTURE_SOURCE_ROWS + 1).all();
+    if (sourceRows.length > MAX_STRUCTURE_SOURCE_ROWS) {
+      throw validationError(`Persisted crawl exceeds the ${MAX_STRUCTURE_SOURCE_ROWS}-page structure limit`);
+    }
+    const children = /* @__PURE__ */ new Map();
+    for (const row of sourceRows) {
+      const childPath = structureChildPath(parentPath, row.path);
+      if (!childPath) continue;
+      const child = children.get(childPath) ?? {
+        url: null,
+        hasPage: false,
+        pageCount: 0,
+        inventoryEligibleCount: 0,
+        fetchedCount: 0
+      };
+      child.pageCount++;
+      if (row.inventoryEligible) child.inventoryEligibleCount++;
+      if (["html", "redirect", "non-html", "fetch-error", "fetched"].includes(row.fetchState)) child.fetchedCount++;
+      if (structureHierarchyPath(row.path) === childPath) {
+        child.hasPage = true;
+        if (child.url == null || row.url > child.url) child.url = row.url;
+      }
+      children.set(childPath, child);
+    }
+    const limit = parseBoundedLimit(request.query.limit, 50, 100);
+    const offset = decodeCursor(request.query.cursor);
+    const sortedChildren = [...children.entries()].sort(([left], [right]) => left.localeCompare(right));
+    const page = sortedChildren.slice(offset, offset + limit + 1);
+    const hasMore = page.length > limit;
+    const visible = hasMore ? page.slice(0, limit) : page;
+    return {
+      project: project.name,
+      hasCrawlData: true,
+      runId: snapshot.runId,
+      parentPath,
+      nextCursor: hasMore ? encodeCursor2(offset + visible.length) : null,
+      children: visible.map(([path8, child]) => ({
+        path: path8,
+        url: child.url,
+        hasPage: child.hasPage,
+        pageCount: child.pageCount,
+        inventoryEligibleCount: child.inventoryEligibleCount,
+        fetchedCount: child.fetchedCount
+      }))
+    };
+  });
+  app.get("/projects/:name/technical-aeo/internal-links", async (request) => {
+    const project = resolveProject(app.db, request.params.name);
+    const target = resolveCrawl(project.id, request.query.runId);
+    if (!target) {
+      if (request.query.runId) throw notFound("Site crawl run", request.query.runId);
+      return { project: project.name, hasCrawlData: false, runId: null, total: 0, nextCursor: null, edges: [] };
+    }
+    const snapshot = target.snapshot;
+    if (!snapshot.detailsAvailable || !snapshot.attemptId) {
+      return { project: project.name, hasCrawlData: true, runId: snapshot.runId, total: 0, nextCursor: null, edges: [] };
+    }
+    const filters = [
+      eq55(siteCrawlEdges.projectId, project.id),
+      eq55(siteCrawlEdges.runId, snapshot.runId),
+      eq55(siteCrawlEdges.attemptId, snapshot.attemptId),
+      eq55(siteCrawlEdges.internal, true)
+    ];
+    if (request.query.sourceUrl) filters.push(eq55(siteCrawlEdges.sourceUrl, request.query.sourceUrl));
+    if (request.query.targetUrl) filters.push(eq55(siteCrawlEdges.targetUrl, request.query.targetUrl));
+    const followable = parseBoolean2(request.query.followable);
+    if (followable != null) filters.push(eq55(siteCrawlEdges.followable, followable));
+    const where = and43(...filters);
+    const total = app.db.select({ value: count() }).from(siteCrawlEdges).where(where).get()?.value ?? 0;
+    const limit = parseBoundedLimit(request.query.limit, 100, 200);
+    const offset = decodeCursor(request.query.cursor);
+    const rows = app.db.select().from(siteCrawlEdges).where(where).orderBy(asc10(siteCrawlEdges.edgeKey)).limit(limit).offset(offset).all();
+    const nextOffset = offset + rows.length;
+    return {
+      project: project.name,
+      hasCrawlData: true,
+      runId: snapshot.runId,
+      total,
+      nextCursor: nextOffset < total ? encodeCursor2(nextOffset) : null,
+      edges: rows.map(mapCrawlEdge)
+    };
+  });
+  app.get("/projects/:name/technical-aeo/internal-links/neighbors", async (request) => {
+    const project = resolveProject(app.db, request.params.name);
+    if (!request.query.nodeKey && !request.query.url) {
+      throw validationError("Provide nodeKey or url");
+    }
+    const target = resolveCrawl(project.id, request.query.runId);
+    if (!target) {
+      if (request.query.runId) throw notFound("Site crawl run", request.query.runId);
+      return {
+        project: project.name,
+        hasCrawlData: false,
+        runId: null,
+        nodeKey: request.query.nodeKey ?? null,
+        url: request.query.url ?? null,
+        inbound: [],
+        outbound: [],
+        inboundTruncated: false,
+        outboundTruncated: false
+      };
+    }
+    const snapshot = target.snapshot;
+    if (!snapshot.detailsAvailable || !snapshot.attemptId) {
+      return {
+        project: project.name,
+        hasCrawlData: true,
+        runId: snapshot.runId,
+        nodeKey: request.query.nodeKey ?? null,
+        url: request.query.url ?? null,
+        inbound: [],
+        outbound: [],
+        inboundTruncated: false,
+        outboundTruncated: false
+      };
+    }
+    const scope = [
+      eq55(siteCrawlEdges.projectId, project.id),
+      eq55(siteCrawlEdges.runId, snapshot.runId),
+      eq55(siteCrawlEdges.attemptId, snapshot.attemptId),
+      eq55(siteCrawlEdges.internal, true)
+    ];
+    const inboundMatch = request.query.nodeKey && request.query.url ? or11(eq55(siteCrawlEdges.targetNodeKey, request.query.nodeKey), eq55(siteCrawlEdges.targetUrl, request.query.url)) : request.query.nodeKey ? eq55(siteCrawlEdges.targetNodeKey, request.query.nodeKey) : eq55(siteCrawlEdges.targetUrl, request.query.url);
+    const outboundMatch = request.query.nodeKey && request.query.url ? or11(eq55(siteCrawlEdges.sourceNodeKey, request.query.nodeKey), eq55(siteCrawlEdges.sourceUrl, request.query.url)) : request.query.nodeKey ? eq55(siteCrawlEdges.sourceNodeKey, request.query.nodeKey) : eq55(siteCrawlEdges.sourceUrl, request.query.url);
+    const limit = parseBoundedLimit(request.query.limit, 50, 100);
+    const inboundRows = app.db.select().from(siteCrawlEdges).where(and43(...scope, inboundMatch)).orderBy(asc10(siteCrawlEdges.edgeKey)).limit(limit + 1).all();
+    const outboundRows = app.db.select().from(siteCrawlEdges).where(and43(...scope, outboundMatch)).orderBy(asc10(siteCrawlEdges.edgeKey)).limit(limit + 1).all();
+    return {
+      project: project.name,
+      hasCrawlData: true,
+      runId: snapshot.runId,
+      nodeKey: request.query.nodeKey ?? null,
+      url: request.query.url ?? null,
+      inbound: inboundRows.slice(0, limit).map(mapCrawlEdge),
+      outbound: outboundRows.slice(0, limit).map(mapCrawlEdge),
+      inboundTruncated: inboundRows.length > limit,
+      outboundTruncated: outboundRows.length > limit
+    };
+  });
+  app.get("/projects/:name/technical-aeo/dead-links", async (request) => {
+    const project = resolveProject(app.db, request.params.name);
+    const target = resolveCrawl(project.id, request.query.runId);
+    const legacyAuditAvailable = hasLegacyAudit(project.id);
+    if (!target) {
+      if (request.query.runId) throw notFound("Site crawl run", request.query.runId);
+      return { project: project.name, runId: null, state: "unavailable", legacyAuditAvailable };
+    }
+    const snapshot = target.snapshot;
+    if (!snapshot.checkDeadLinks || snapshot.deadLinkState === "disabled") {
+      return { project: project.name, runId: snapshot.runId, state: "disabled", checkDeadLinks: false };
+    }
+    if (snapshot.deadLinkState === "unavailable") {
+      return { project: project.name, runId: snapshot.runId, state: "unavailable", legacyAuditAvailable };
+    }
+    if (!snapshot.attemptId) {
+      return { project: project.name, runId: snapshot.runId, state: "partial", checkDeadLinks: true, checked: snapshot.deadLinksChecked, found: snapshot.deadLinksFound, total: 0, nextCursor: null, deadLinks: [] };
+    }
+    const where = and43(
+      eq55(siteCrawlFindings.projectId, project.id),
+      eq55(siteCrawlFindings.runId, snapshot.runId),
+      eq55(siteCrawlFindings.attemptId, snapshot.attemptId),
+      eq55(siteCrawlFindings.findingType, "dead-link")
+    );
+    const total = app.db.select({ value: count() }).from(siteCrawlFindings).where(where).get()?.value ?? 0;
+    const limit = parseBoundedLimit(request.query.limit, 100, 200);
+    const offset = decodeCursor(request.query.cursor);
+    const rows = app.db.select().from(siteCrawlFindings).where(where).orderBy(asc10(siteCrawlFindings.findingKey)).limit(limit).offset(offset).all();
+    const nextOffset = offset + rows.length;
+    const deadLinks = rows.map((row) => ({
+      findingKey: row.findingKey,
+      severity: row.severity,
+      sourceNodeKey: row.sourceNodeKey,
+      sourceUrl: row.sourceUrl,
+      targetNodeKey: row.targetNodeKey,
+      targetUrl: row.targetUrl,
+      evidence: row.evidence
+    }));
+    const state = snapshot.deadLinkState === "complete" ? "complete" : "partial";
+    return {
+      project: project.name,
+      runId: snapshot.runId,
+      state,
+      checkDeadLinks: true,
+      checked: snapshot.deadLinksChecked,
+      found: snapshot.deadLinksFound,
+      total,
+      nextCursor: nextOffset < total ? encodeCursor2(nextOffset) : null,
+      deadLinks
+    };
+  });
   app.post("/projects/:name/technical-aeo/runs", async (request) => {
     const project = resolveProject(app.db, request.params.name);
     const parsed = siteAuditRunRequestSchema.safeParse(request.body ?? {});
     if (!parsed.success) {
       throw validationError(parsed.error.issues[0]?.message ?? "Invalid site-audit request");
     }
-    const existing = app.db.select({ id: runs.id, status: runs.status }).from(runs).where(and43(
-      eq55(runs.projectId, project.id),
-      eq55(runs.kind, RunKinds["site-audit"]),
-      inArray18(runs.status, [RunStatuses.queued, RunStatuses.running])
-    )).get();
-    if (existing) {
-      return { runId: existing.id, status: existing.status };
+    const effectiveRequest = normalizeSiteAuditRunRequest(parsed.data);
+    const identityKey = siteAuditRequestIdentity(effectiveRequest);
+    const result = app.db.transaction((tx) => {
+      const existing = tx.select({
+        id: runs.id,
+        status: runs.status,
+        identityKey: siteCrawlRunRequests.identityKey,
+        effectiveOptions: siteCrawlRunRequests.effectiveOptions
+      }).from(runs).leftJoin(siteCrawlRunRequests, and43(
+        eq55(siteCrawlRunRequests.projectId, runs.projectId),
+        eq55(siteCrawlRunRequests.runId, runs.id)
+      )).where(and43(
+        eq55(runs.projectId, project.id),
+        eq55(runs.kind, RunKinds["site-audit"]),
+        inArray19(runs.status, [RunStatuses.queued, RunStatuses.running])
+      )).get();
+      if (existing) {
+        if (existing.identityKey === identityKey) {
+          return { runId: existing.id, status: existing.status, created: false };
+        }
+        throw operationInProgress(
+          `A Technical AEO crawl is already ${existing.status} with different request options.`,
+          {
+            activeRunId: existing.id,
+            activeStatus: existing.status,
+            activeOptions: existing.effectiveOptions,
+            requestedOptions: effectiveRequest
+          }
+        );
+      }
+      const now = (/* @__PURE__ */ new Date()).toISOString();
+      const runId = crypto43.randomUUID();
+      tx.insert(runs).values({
+        id: runId,
+        projectId: project.id,
+        kind: RunKinds["site-audit"],
+        status: RunStatuses.queued,
+        trigger: RunTriggers.manual,
+        createdAt: now
+      }).run();
+      tx.insert(siteCrawlRunRequests).values({
+        runId,
+        projectId: project.id,
+        identityKey,
+        effectiveOptions: effectiveRequest,
+        createdAt: now
+      }).run();
+      return { runId, status: RunStatuses.queued, created: true };
+    });
+    if (result.created) {
+      opts.onSiteAuditRequested?.(result.runId, project.id, {
+        sitemapUrl: effectiveRequest.sitemapUrl ?? void 0,
+        limit: parsed.data.limit,
+        maxPages: effectiveRequest.maxPages,
+        maxEdges: effectiveRequest.maxEdges,
+        maxDepth: effectiveRequest.maxDepth ?? void 0,
+        checkDeadLinks: effectiveRequest.checkDeadLinks
+      });
     }
-    const now = (/* @__PURE__ */ new Date()).toISOString();
-    const runId = crypto43.randomUUID();
-    app.db.insert(runs).values({
-      id: runId,
-      projectId: project.id,
-      kind: RunKinds["site-audit"],
-      status: RunStatuses.queued,
-      trigger: RunTriggers.manual,
-      createdAt: now
-    }).run();
-    opts.onSiteAuditRequested?.(runId, project.id, {
-      sitemapUrl: parsed.data.sitemapUrl,
-      limit: parsed.data.limit
-    });
-    return { runId, status: RunStatuses.queued };
+    return { runId: result.runId, status: result.status };
   });
 }
 
@@ -55113,6 +56126,7 @@ async function apiRoutes(app, opts) {
     await api.register(competitorRoutes);
     await api.register(runRoutes, {
       onRunCreated: opts.onRunCreated,
+      onRunCancelled: opts.onRunCancelled,
       validProviderNames: opts.providerAdapters?.map((a) => a.name),
       getRunnableProviderNames: opts.getRunnableProviderNames,
       getEffectiveProviderModels: opts.getEffectiveProviderModels
@@ -55634,7 +56648,7 @@ var IntelligenceService = class {
     const recentRuns = this.db.select().from(runs).where(
       and45(
         eq57(runs.projectId, projectId),
-        or11(eq57(runs.status, "completed"), eq57(runs.status, "partial")),
+        or12(eq57(runs.status, "completed"), eq57(runs.status, "partial")),
         // Defensive: RunCoordinator already skips probes before this is
         // called, but if a future call site invokes analyzeAndPersist
         // directly for a probe, probes still must not pollute the
@@ -55937,7 +56951,7 @@ var IntelligenceService = class {
     const allRuns = this.db.select().from(runs).where(
       and45(
         eq57(runs.projectId, project.id),
-        or11(eq57(runs.status, "completed"), eq57(runs.status, "partial")),
+        or12(eq57(runs.status, "completed"), eq57(runs.status, "partial")),
         // Backfill must not replay probe runs as if they were real sweeps.
         ne7(runs.trigger, RunTriggers.probe)
       )
@@ -55970,7 +56984,7 @@ var IntelligenceService = class {
     let wouldDeleteTotal = 0;
     const existingByRunId = /* @__PURE__ */ new Map();
     if (isDryRun && targetRuns.length > 0) {
-      const rows = this.db.select({ runId: insights.runId }).from(insights).where(inArray19(insights.runId, targetRuns.map((r) => r.id))).all();
+      const rows = this.db.select({ runId: insights.runId }).from(insights).where(inArray20(insights.runId, targetRuns.map((r) => r.id))).all();
       for (const r of rows) {
         if (r.runId == null) continue;
         existingByRunId.set(r.runId, (existingByRunId.get(r.runId) ?? 0) + 1);
@@ -56115,7 +57129,7 @@ var IntelligenceService = class {
       and45(
         eq57(runs.projectId, projectId),
         eq57(runs.kind, RunKinds["answer-visibility"]),
-        or11(eq57(runs.status, "completed"), eq57(runs.status, "partial")),
+        or12(eq57(runs.status, "completed"), eq57(runs.status, "partial")),
         // Defensive — see top of file.
         ne7(runs.trigger, RunTriggers.probe)
       )
@@ -56135,7 +57149,7 @@ var IntelligenceService = class {
     const haveHistory = recentRunIds.length > 0;
     const priorRegressionsByPair = /* @__PURE__ */ new Map();
     if (haveHistory) {
-      const priorRows = this.db.select({ query: insights.query, provider: insights.provider, runId: insights.runId }).from(insights).where(and45(eq57(insights.type, "regression"), inArray19(insights.runId, recentRunIds))).all();
+      const priorRows = this.db.select({ query: insights.query, provider: insights.provider, runId: insights.runId }).from(insights).where(and45(eq57(insights.type, "regression"), inArray20(insights.runId, recentRunIds))).all();
       const regressionGroups = /* @__PURE__ */ new Map();
       for (const row of priorRows) {
         if (!row.runId) continue;
@@ -56241,6 +57255,13 @@ export {
   gscCoverageSnapshots,
   siteAuditSnapshots,
   siteAuditPages,
+  siteCrawlRunRequests,
+  siteCrawlAttempts,
+  siteCrawlSnapshots,
+  siteCrawlPages,
+  siteCrawlEdges,
+  siteCrawlFindings,
+  siteCrawlEventReceipts,
   bingCoverageSnapshots,
   bingUrlInspections,
   gaTrafficSnapshots,