mcpscraper-sdk 0.40.1 → 0.42.0

package/README.md

@@ -45,7 +45,7 @@ Current Google search pricing is 60 Credits per SERP search and 400 Credits plus
 
 ## API surface
 
-`client.tools` is the generated, typed 377-tool MCP surface from `contracts/mcp.tools.json`, including provider-neutral X-Ray surveys, attribution-impact reporting, and truthful evidence-status operations.
+`client.tools` is the generated, typed 375-tool MCP surface from `contracts/mcp.tools.json`, including provider-neutral X-Ray surveys, attribution-impact reporting, and truthful evidence-status operations.
 
 For multimodal results such as `meta_ad_creative_media`, call `client.tools.callToolResult(...)` to preserve native MCP image/audio/resource blocks. `callTool(...)` remains backward-compatible and returns the parsed structured or text value.
 
@@ -89,6 +89,13 @@ Integrations are included with an active Starter plan or higher: OAuth connect/r
 
 Core operations are flat on the client: `startHarvest`, `searchSerp`, `harvestPaa`, `extractUrl`, `mapSiteUrls`, `extractSite`, `auditSite`, `getExtractSiteStatus`, `archiveRead`, `listJobs`, `getJob`, `getHistory`, `getLedger`.
 
+For a durable PAA harvest, `client.tools.other.harvestPaaStart` accepts
+`pages: 1 | 2` and defaults to one. Two pages add page-two organic results when
+available while expanding only the preserved page-one PAA graph. Poll the same
+`jobId` with `harvestPaaStatus`; its `progress.pagination` and
+`result.pagination` are nullable for older saved jobs. The complete workflow is
+in [`../../examples/paa-two-page.mjs`](../../examples/paa-two-page.mjs).
+
 ### Durable acquisition
 
 Long-running agents can start SERP/PAA work once and poll the provider-owned job
@@ -170,7 +177,7 @@ const hits = await client.memoryTools.memory.search({ query: 'competitor pricing
 const vaults = await client.memoryTools.vaults.listVaults({})
 ```
 
-Use `mcpscraper-memory-sdk`'s own `MemoryClient` instead if you already have a dedicated `mk_...` memory key and want to talk to `memory.mcpscraper.dev` directly.
+Use `mcpscraper-memory-sdk`'s `MemoryClient` when you prefer its Memory-first namespaces; it uses the same MCP Scraper key and root endpoint.
 
 ## Regenerating types
 

package/dist/index.d.cts

@@ -1072,7 +1072,7 @@ interface paths$1 {
         get?: never;
         put?: never;
         /**
-         * Call one memory.mcpscraper.dev tool by name, using this account's mcpscraper.dev API key
+         * Call one Memory tool by name through mcpscraper.dev using this account's API key
          * @description Generic dispatch: names one of the 116 tools documented in this repo's
          *     `contracts/memory.tools.json` and forwards `args` to it, using a memory identity
          *     auto-provisioned for the calling mcpscraper.dev account. The response shape depends

package/dist/index.d.ts

@@ -1072,7 +1072,7 @@ interface paths$1 {
         get?: never;
         put?: never;
         /**
-         * Call one memory.mcpscraper.dev tool by name, using this account's mcpscraper.dev API key
+         * Call one Memory tool by name through mcpscraper.dev using this account's API key
          * @description Generic dispatch: names one of the 116 tools documented in this repo's
          *     `contracts/memory.tools.json` and forwards `args` to it, using a memory identity
          *     auto-provisioned for the calling mcpscraper.dev account. The response shape depends

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "mcpscraper-sdk",
-  "version": "0.40.1",
-  "description": "Official Node.js client for all 377 mcpscraper.dev MCP tools and the REST API",
+  "version": "0.42.0",
+  "description": "Official Node.js client for all 375 mcpscraper.dev MCP tools and the REST API",
   "license": "MIT",
   "type": "module",
   "main": "./dist/index.cjs",
@@ -32,7 +32,7 @@
     "seo-api"
   ],
   "dependencies": {
-    "mcpscraper-memory-sdk": "^0.40.1"
+    "mcpscraper-memory-sdk": "^0.42.0"
   },
   "devDependencies": {
     "openapi-typescript": "^7.13.0"