mcp-scraper 0.48.0 → 0.50.0

package/dist/{chunk-PTABZABJ.js → chunk-44LNKZTE.js}

@@ -18,10 +18,10 @@ import {
   browserServiceProfileSaveChanges,
   recordVendorUsage,
   vendorCostUsd
-} from "./chunk-PODASLGT.js";
+} from "./chunk-QXAY44SA.js";
 import {
   PACKAGE_VERSION
-} from "./chunk-IAXQIHNS.js";
+} from "./chunk-C7NN57WH.js";
 import {
   MC_PER_CREDIT
 } from "./chunk-SOEYDWJU.js";
@@ -33,6 +33,7 @@ import {
 
 // src/harvest-timeout.ts
 var VERCEL_FUNCTION_MAX_MS = 3e5;
+var DEFAULT_TOOL_CLIENT_TIMEOUT_MS = 3e5;
 var CLIENT_OVER_SERVER_MARGIN_MS = 15e3;
 function harvestTimeoutBudget(maxQuestions, serpOnly = false) {
   const requested = Number.isFinite(maxQuestions) && maxQuestions > 0 ? Math.trunc(maxQuestions) : 30;
@@ -360,6 +361,8 @@ seam is noted so you can chain them.
   JSON-LD, media, citations, source metadata, related entities, trails, and saved account filters.
 - Read one published entity -> **commons_get_entity**. It returns the Wikidata-style backend fields,
   JSON-LD, source provenance, media, ledger, and the Wikipedia-style page projection.
+- Read its machine-verifiable governed relationships -> **commons_get_entity_linkset**. It returns the
+  RFC 9264 linkset projection; do not infer claim status from loose See Also labels.
 - List unresolved See Also concepts -> **commons_list_needs_links**. Use it to find concepts that appear
   in public See Also sections but do not yet have a Commons entity ID or /wiki/ slug.
 - Plan a new or edited entity -> **commons_prepare_entity**. This is the Commons equivalent of
@@ -5329,6 +5332,9 @@ var CommonsGetEntityInputSchema = {
   idOrSlug: z4.string().trim().min(1).max(180).describe("Transparent Public Wiki entity id such as TPW-Q... or a public /wiki/ slug."),
   includeWikiPage: z4.boolean().default(true).describe("Include the Wikipedia-style page projection used by transparent-commons.cc/wiki/.")
 };
+var CommonsGetEntityLinksetInputSchema = {
+  idOrSlug: z4.string().trim().min(1).max(180).describe("Transparent Commons TPW-Q entity id or /wiki/ slug whose governed RFC 9264 linkset should be read.")
+};
 var CommonsFeaturedImageInputSchema = z4.object({
   url: z4.string().url().describe("Required public image URL for a publishable entity. Use extract_url includeFeaturedImage or preserved media when available."),
   alt: z4.string().trim().max(500).optional(),
@@ -5381,6 +5387,27 @@ var CommonsRelatedLinkInputSchema = z4.object({
   imageUrl: z4.string().trim().max(2e3).optional().describe("Optional preview image for hover cards when available."),
   needsLink: z4.boolean().default(false).describe("Set true when the concept belongs in See Also but no existing Commons entity/page exists yet. The backend exposes these records through commons_list_needs_links.")
 }).strict();
+var CommonsClaimInputSchema = z4.object({
+  predicate: z4.string().trim().min(1).max(500).describe("Registered link relation such as related, describedby, license, or type; alternatively an absolute HTTPS relation URI from a published vocabulary."),
+  predicateLabel: z4.string().trim().max(160).optional().describe("Human-readable label for display and review. It does not replace the machine predicate."),
+  objectEntityId: z4.string().trim().regex(/^TPW-Q[A-Za-z0-9._-]+$/).max(80).optional().describe("Existing Transparent Commons entity target. Use exactly one of objectEntityId or objectUrl."),
+  objectUrl: z4.string().url().refine((value) => value.startsWith("https://"), "objectUrl must use HTTPS.").optional().describe("External or third-party target. Use exactly one of objectUrl or objectEntityId."),
+  title: z4.string().trim().max(500).optional(),
+  mediaType: z4.string().trim().max(160).optional().describe("Optional RFC link target media type."),
+  hreflang: z4.array(z4.string().trim().min(1).max(80)).max(20).optional().describe("Optional repeatable RFC link target language hints."),
+  qualifiers: z4.record(z4.string().max(120), z4.union([z4.string(), z4.number(), z4.boolean(), z4.null()])).optional(),
+  references: z4.array(CommonsCitationInputSchema).max(50).optional().describe("Evidence supporting the relationship assertion."),
+  rank: z4.enum(["preferred", "normal", "deprecated"]).default("normal"),
+  confidence: z4.number().min(0).max(1).default(1),
+  confidenceReason: z4.string().trim().max(1e3).optional(),
+  disputeState: z4.enum(["undisputed", "attributed", "disputed", "contested", "needs_review", "deprecated"]).default("undisputed"),
+  validFrom: z4.string().trim().max(80).optional(),
+  validTo: z4.string().trim().max(80).optional(),
+  asOf: z4.string().trim().max(80).optional(),
+  sourceFamily: z4.enum(["first_party", "partner", "independent", "community", "unknown"]).default("unknown")
+}).strict().refine((claim) => Boolean(claim.objectEntityId) !== Boolean(claim.objectUrl), {
+  message: "Provide exactly one of objectEntityId or objectUrl."
+});
 var CommonsPrepareEntityInputSchema = {
   title: z4.string().trim().min(1).max(180).describe("Candidate public page title. Used to derive slug, choose a profile, and search duplicate entities."),
   description: z4.string().trim().max(600).optional().describe("Optional candidate summary. Include when known so the prepare pass can route and validate the article shape."),
@@ -5420,6 +5447,7 @@ var CommonsSubmitEntityInputSchema = {
     url: z4.string().url().optional(),
     description: z4.string().trim().max(300).optional()
   }).strict()).max(200).optional().describe("Related concepts, including trails. Existing entities should use entityId."),
+  claims: z4.array(CommonsClaimInputSchema).max(300).optional().describe("Governed relationship assertions. Supplying claims replaces the accepted claim set for this entity; omitting claims preserves it. Only approved, active, non-disputed claims project into the RFC 9264 linkset."),
   bodyMarkdown: z4.string().max(5e5).optional().describe("Structured encyclopedia body in neutral Markdown, not a raw scrape dump or blog essay. Use H2 for entity-profile sections and H3/H4/H5 for subtopics; every heading appears in the public page menu. Call commons_prepare_entity first for the live entity profile and commons_validate_entity before submit. Profile sections are adaptive: include History, Pricing, Reception, Timeline, Classification, or similar sections only when source evidence supports them; omit unsupported sections instead of adding empty/filler headings. Apply NPOV, verifiability, and no-original-research rules; self-published sources can support only uncontroversial source-owned facts."),
   contentSections: z4.array(z4.object({
     id: z4.string().trim().max(180).optional(),
@@ -8631,6 +8659,19 @@ function registerPaaExtractorMcpTools(server, executor, options = {}) {
       openWorldHint: false
     }
   }, async (input) => executor.commonsGetEntity(input));
+  server.registerTool("commons_get_entity_linkset", {
+    title: "Transparent Commons RFC 9264 Linkset",
+    description: "Read the governed RFC 9264 application/linkset+json projection for one published Transparent Commons entity. The anchor is the canonical /wiki/ page; only approved, active, non-disputed claims are included, while third-party targets remain explicit HTTPS links.",
+    inputSchema: CommonsGetEntityLinksetInputSchema,
+    outputSchema: recordOutputSchema("commons_get_entity_linkset", CommonsGenericOutputSchema),
+    annotations: {
+      title: "Transparent Commons RFC 9264 Linkset",
+      readOnlyHint: true,
+      destructiveHint: false,
+      idempotentHint: true,
+      openWorldHint: false
+    }
+  }, async (input) => executor.commonsGetEntityLinkset(input));
   server.registerTool("commons_list_needs_links", {
     title: "Transparent Commons Needs-Link Backlog",
     description: "List unresolved See Also concepts from published Transparent Commons pages. Use this after search/lookup when an agent needs to grow the graph: each result names a concept that appears in See Also but does not yet have a resolved Commons entityId or /wiki/ slug. This is platform Commons graph data, not personal Memory storage.",
@@ -9098,6 +9139,29 @@ function registerPaaExtractorMcpTools(server, executor, options = {}) {
 
 // src/mcp/http-mcp-tool-executor.ts
 import { createHash as createHash4, randomUUID as randomUUID2 } from "crypto";
+function unclassifiedTransportFailure(path, err) {
+  const name = err instanceof Error ? err.name : typeof err;
+  const detail = err instanceof Error ? err.message : String(err);
+  console.error(JSON.stringify({
+    event: "mcp_executor_transport_failed",
+    path,
+    error_name: name,
+    error_message: sanitizeVendorName(detail).slice(0, 400)
+  }));
+  return {
+    content: [{
+      type: "text",
+      text: JSON.stringify({
+        error: "service_unavailable",
+        error_type: name,
+        retryable: true,
+        path,
+        message: publicErrorMessage("service_unavailable")
+      })
+    }],
+    isError: true
+  };
+}
 function youtubeVideoIdFromUrl(url) {
   if (!url) return null;
   try {
@@ -9198,7 +9262,7 @@ var HttpMcpToolExecutor = class {
     const rawOverride = process.env.MCP_SCRAPER_HTTP_TIMEOUT_MS;
     const parsedOverride = rawOverride === void 0 ? NaN : Number(rawOverride);
     this.httpTimeoutOverrideMs = Number.isFinite(parsedOverride) && parsedOverride > 0 ? parsedOverride : null;
-    this.timeoutMs = this.httpTimeoutOverrideMs ?? 11e4;
+    this.timeoutMs = this.httpTimeoutOverrideMs ?? DEFAULT_TOOL_CLIENT_TIMEOUT_MS;
     const configuredSerpIntelligenceTimeoutMs = Number(process.env.MCP_SCRAPER_SERP_INTELLIGENCE_HTTP_TIMEOUT_MS ?? this.timeoutMs);
     this.serpIntelligenceTimeoutMs = Number.isFinite(configuredSerpIntelligenceTimeoutMs) && configuredSerpIntelligenceTimeoutMs > 0 ? configuredSerpIntelligenceTimeoutMs : this.timeoutMs;
   }
@@ -9237,7 +9301,7 @@ var HttpMcpToolExecutor = class {
           isError: true
         };
       }
-      return { content: [{ type: "text", text: publicErrorMessage("service_unavailable") }], isError: true };
+      return unclassifiedTransportFailure(path, err);
     }
   }
   async callConnectedMutation(path, body, idempotencyKey, timeoutMs = this.timeoutMs) {
@@ -9271,8 +9335,8 @@ var HttpMcpToolExecutor = class {
         return { content: [{ type: "text", text: JSON.stringify(httpErrorPayload(path, res, data)) }], isError: true };
       }
       return { content: [{ type: "text", text: JSON.stringify(data) }] };
-    } catch {
-      return { content: [{ type: "text", text: publicErrorMessage("service_unavailable") }], isError: true };
+    } catch (err) {
+      return unclassifiedTransportFailure(path, err);
     }
   }
   async getTextArtifact(path, maxBytes, timeoutMs = this.timeoutMs) {
@@ -9309,7 +9373,7 @@ var HttpMcpToolExecutor = class {
         }]
       };
     } catch (err) {
-      return { content: [{ type: "text", text: publicErrorMessage("service_unavailable") }], isError: true };
+      return unclassifiedTransportFailure(path, err);
     }
   }
   harvestPaa(input) {
@@ -9666,6 +9730,9 @@ var HttpMcpToolExecutor = class {
       return result;
     }
   }
+  commonsGetEntityLinkset(input) {
+    return this.getJson(`/commons/entities/${encodeURIComponent(input.idOrSlug)}/linkset?envelope=true`);
+  }
   commonsPrepareEntity(input) {
     return this.call("/commons/entities/prepare", input);
   }
@@ -14401,6 +14468,7 @@ export {
   EDITORIAL_READING_ROOM_TEMPLATE_PRESET,
   PERSONAL_AUTHORITY_TEMPLATE_PRESET,
   NEWSROOM_PUBLISHER_TEMPLATE_PRESET,
+  CommonsClaimInputSchema,
   CreateEditorialReadingRoomInputSchema,
   RenewEditorialReadingRoomDownloadInputSchema,
   CommonsPreparePublicationInputSchema,
@@ -14423,4 +14491,4 @@ export {
   ScheduledResultsMcpExecutor,
   registerScheduledResultsMcpTools
 };
-//# sourceMappingURL=chunk-PTABZABJ.js.map
+//# sourceMappingURL=chunk-44LNKZTE.js.map