fix(ci): fix the monorepo

.dockerignore

@@ -0,0 +1,14 @@
+# Actor images build from the monorepo root (dockerContextDir in .actor/actor.json).
+# Keep the context lean: ship sources + manifests + the pnpm lockfile only.
+node_modules
+**/node_modules
+.git
+.turbo
+**/dist
+**/storage
+apify_storage
+crawlee_storage
+*.log
+.DS_Store
+.idea
+.vscode

.github/workflows/on-pull-request.yml

@@ -13,6 +13,7 @@ jobs:
         name: Build & Test
         if: (!contains(github.event.head_commit.message, '[skip ci]') && !contains(github.event.head_commit.message, 'docs:'))
         runs-on: ubuntu-22.04
+        timeout-minutes: 30
 
         steps:
             - name: Checkout repository
@@ -38,6 +39,7 @@ jobs:
     lint:
         name: Lint
         runs-on: ubuntu-22.04
+        timeout-minutes: 30
         steps:
             - name: Checkout repository
               uses: actions/checkout@v6

.github/workflows/release-generic-actors.yaml

@@ -1,5 +1,4 @@
 name: Build generic Actors
-description: Build generic Actors and push them to Apify
 
 on:
     workflow_dispatch:
@@ -34,6 +33,11 @@ on:
                 type: boolean
                 required: false
                 default: true
+            sitemap-scraper:
+                description: apify/sitemap-extractor
+                type: boolean
+                required: false
+                default: true
             build-channel:
                 description: Build channel
                 type: choice
@@ -60,43 +64,68 @@ jobs:
             matrix:
                 settings:
                     - actor: web-scraper
+                      apify-actor: apify/web-scraper
                       stable-version: '3.0'
                       stable-build-tag: version-3
                       development-version: '0.0'
                       development-build-tag: development
                       should-build: ${{ github.event.inputs.web-scraper }}
                     - actor: cheerio-scraper
+                      apify-actor: apify/cheerio-scraper
                       stable-version: '3.0'
                       stable-build-tag: version-3
                       development-version: '0.0'
                       development-build-tag: development
                       should-build: ${{ github.event.inputs.cheerio-scraper }}
                     - actor: playwright-scraper
+                      apify-actor: apify/playwright-scraper
                       stable-version: '1.0'
                       stable-build-tag: version-1
                       development-version: '0.0'
                       development-build-tag: development
                       should-build: ${{ github.event.inputs.playwright-scraper }}
                     - actor: puppeteer-scraper
+                      apify-actor: apify/puppeteer-scraper
                       stable-version: '3.0'
                       stable-build-tag: version-3
                       development-version: '0.0'
                       development-build-tag: development
                       should-build: ${{ github.event.inputs.puppeteer-scraper }}
                     - actor: jsdom-scraper
+                      apify-actor: apify/jsdom-scraper
                       stable-version: '0.2'
                       stable-build-tag: version-0
                       development-version: '0.0'
                       development-build-tag: development
                       should-build: ${{ github.event.inputs.jsdom-scraper }}
                     - actor: camoufox-scraper
+                      apify-actor: apify/camoufox-scraper
                       stable-version: '3.0'
                       stable-build-tag: version-3
                       development-version: '0.0'
                       development-build-tag: development
                       should-build: ${{ github.event.inputs.camoufox-scraper }}
+                    - actor: sitemap-scraper
+                      apify-actor: apify/sitemap-extractor
+                      stable-version: '0.1'
+                      stable-build-tag: latest
+                      development-version: '0.0'
+                      development-build-tag: development
+                      should-build: ${{ github.event.inputs.sitemap-scraper }}
         steps:
-            - uses: actions/checkout@v6
+            - name: Check out current SHA
+              if: matrix.settings.should-build == 'true'
+              uses: actions/checkout@v6
+
+            - name: Set up Node.js
+              if: matrix.settings.should-build == 'true'
+              uses: actions/setup-node@v6
+              with:
+                  node-version: 24
+
+            - name: Install pnpm and dependencies
+              if: matrix.settings.should-build == 'true'
+              uses: apify/actions/pnpm-install@v1.1.2
 
             - name: Log matrix
               run: |
@@ -108,21 +137,24 @@ jobs:
               if: matrix.settings.should-build == 'true'
               run: |
                   if [ "${{ github.event.inputs.build-channel }}" = "stable" ]; then
-                      echo "version=${{ matrix.settings.stable-version }}" >> $GITHUB_ENV
-                      echo "build-tag=${{ matrix.settings.stable-build-tag }}" >> $GITHUB_ENV
+                      echo "version=${{ matrix.settings.stable-version }}" >> "$GITHUB_ENV"
+                      echo "build-tag=${{ matrix.settings.stable-build-tag }}" >> "$GITHUB_ENV"
                   elif [ "${{ github.event.inputs.build-channel }}" = "development" ]; then
-                      echo "version=${{ matrix.settings.development-version }}" >> $GITHUB_ENV
-                      echo "build-tag=${{ matrix.settings.development-build-tag }}" >> $GITHUB_ENV
+                      echo "version=${{ matrix.settings.development-version }}" >> "$GITHUB_ENV"
+                      echo "build-tag=${{ matrix.settings.development-build-tag }}" >> "$GITHUB_ENV"
                   else
-                      echo "version=${{ github.event.inputs.version }}" >> $GITHUB_ENV
-                      echo "build-tag=${{ github.event.inputs.build-tag }}" >> $GITHUB_ENV
+                      echo "version=${{ github.event.inputs.version }}" >> "$GITHUB_ENV"
+                      echo "build-tag=${{ github.event.inputs.build-tag }}" >> "$GITHUB_ENV"
                   fi
 
-            - name: Build ${{ matrix.settings.actor }}
-              uses: apify/push-actor-action@master
+            - name: Build ${{ matrix.settings.actor }} on Apify (Git source)
               if: matrix.settings.should-build == 'true'
-              with:
-                  token: ${{ secrets.APIFY_ACTOR_BUILD_TOKEN }}
-                  build-tag: ${{ env.build-tag }}
-                  version: ${{ env.version }}
-                  working-directory: packages/actor-scraper/${{ matrix.settings.actor }}
+              env:
+                  APIFY_TOKEN: ${{ secrets.APIFY_ACTOR_BUILD_TOKEN }}
+                  APIFY_ACTOR: ${{ matrix.settings.apify-actor }}
+                  BUILD_VERSION: ${{ env.version }}
+                  BUILD_TAG: ${{ env.build-tag }}
+                  APIFY_RELEASE_BUILD_TIMEOUT_SECS: "900"
+              run: |
+                  set -euo pipefail
+                  node scripts/trigger-apify-build.mjs

.github/workflows/test-e2e.yaml

@@ -13,6 +13,7 @@ jobs:
     build_and_test:
         name: Build & Test
         runs-on: ubuntu-22.04
+        timeout-minutes: 60
 
         steps:
             - name: Cancel Workflow Action
@@ -40,6 +41,11 @@ jobs:
             - name: Install pnpm and dependencies
               uses: apify/actions/pnpm-install@v1.1.2
 
+            - name: Install browsers
+              run: |
+                  pnpm exec puppeteer browsers install chrome
+                  pnpm exec playwright install chromium
+
             - name: Build
               run: pnpm ci:build
 

CONTRIBUTING.md

@@ -14,14 +14,14 @@ Please provide steps to reproduce if you found a bug or ideally fork the reposit
 
 Before you submit your pull request, consider the following guidelines:
 
-- Search [GitHub](https://github.com/apify/apify-sdk-js/pulls) for an open or closed PR that relates to your submission. You don't want to duplicate effort.
+- Search [GitHub](https://github.com/apify/actor-scraper/pulls) for an open or closed PR that relates to your submission. You don't want to duplicate effort.
 
-- Fork the project and install NPM dependencies.
+- Fork the project and install dependencies with pnpm.
 
 - Run tests before you start working, to be sure they all pass and your setup is working correctly:
 
     ```sh
-    npm run test
+    pnpm test
     ```
 
 - Be sure to **include appropriate test cases**. Tests help make it clear what the PR is fixing and also make sure the changes won't break over time.

package.json

@@ -1,7 +1,7 @@
 {
-    "name": "apify-sdk-js",
+    "name": "actor-scraper",
     "private": true,
-    "description": "Apify SDK monorepo",
+    "description": "Apify generic scrapers monorepo",
     "keywords": [
         "apify",
         "headless",
@@ -23,12 +23,12 @@
     "license": "Apache-2.0",
     "repository": {
         "type": "git",
-        "url": "git+https://github.com/apify/apify-ts"
+        "url": "git+https://github.com/apify/actor-scraper.git"
     },
     "bugs": {
-        "url": "https://github.com/apify/apify-ts/issues"
+        "url": "https://github.com/apify/actor-scraper/issues"
     },
-    "homepage": "https://sdk.apify.com",
+    "homepage": "https://github.com/apify/actor-scraper",
     "scripts": {
         "prepare": "husky",
         "prepublishOnly": "turbo run copy",
@@ -61,7 +61,7 @@
         "@apify/tsconfig": "^0.1.2",
         "@commitlint/config-conventional": "^20.0.0",
         "@isaacs/brace-expansion": "^5.0.1",
-        "@playwright/browser-chromium": "^1.46.0",
+        "@playwright/browser-chromium": "^1.61.0",
         "@types/content-type": "^1.1.8",
         "@types/fs-extra": "^11.0.4",
         "@types/node": "^24.0.0",
@@ -81,8 +81,8 @@
         "oxfmt": "0.46.0",
         "oxlint": "1.62.0",
         "oxlint-tsgolint": "0.22.0",
-        "playwright": "^1.46.0",
-        "puppeteer": "^24.0.0",
+        "playwright": "^1.61.0",
+        "puppeteer": "25.2.0",
         "rimraf": "^6.0.1",
         "tsx": "^4.16.5",
         "turbo": "2.9.1",

packages/actor-scraper/camoufox-scraper/.actor/actor.json

@@ -3,6 +3,8 @@
     "name": "camoufox-scraper",
     "version": "0.1",
     "buildTag": "latest",
+    "dockerContextDir": "../../../..",
+    "dockerfile": "../Dockerfile",
     "storages": {
         "dataset": {
             "actorSpecification": 1,

packages/actor-scraper/camoufox-scraper/Dockerfile

@@ -1,31 +1,45 @@
-FROM apify/actor-node-playwright-camoufox:22 AS builder
+# Build context is the monorepo root (dockerContextDir in .actor/actor.json), so the build uses the
+# root pnpm-lock.yaml for a deterministic install.
+#
+# The base is pinned by digest (reproducible Camoufox binary + OS deps). The actor pins playwright and
+# camoufox-js to the versions this base ships (24-1.59.1 -> playwright 1.59.1, camoufox-js 0.11.1) and
+# bundles them, so the bundled drivers match the Camoufox binary they drive. The binary itself lives in
+# ~/.cache/camoufox (outside node_modules) and stays in the base. Tag 24-1.59.1 is the newest Camoufox
+# base that survives page JS errors (24-1.60.0 crashes the process on them).
+FROM apify/actor-node-playwright-camoufox:24-1.59.1@sha256:eadc96fa9492284eb45ef70b6b91c841fae7f142d25a22ae2887a21bb78b3469 AS builder
 
-COPY --chown=myuser package*.json ./
+# The browser base runs as `myuser`; the build stage needs root for corepack + a writable workdir.
+USER root
+WORKDIR /app
 
-RUN npm install --include=dev --audit=false
+RUN corepack enable
 
-COPY --chown=myuser . ./
+# Browsers ship with the base image; never let an npm postinstall download them.
+ENV PUPPETEER_SKIP_DOWNLOAD=true \
+    PLAYWRIGHT_SKIP_BROWSER_DOWNLOAD=1
 
-RUN npm run build
+# Whole workspace (root .dockerignore keeps node_modules/.git/dist out of the context).
+COPY . ./
 
-FROM apify/actor-node-playwright-camoufox:22
+# Deterministic install (frozen lockfile, honors minimumReleaseAge) of the actor + its workspace deps.
+RUN pnpm install --frozen-lockfile --filter actor-camoufox-scraper...
 
-COPY --from=builder --chown=myuser /home/myuser/dist ./dist
+# Build the actor and its workspace dependency @apify/scraper-tools.
+RUN pnpm --filter actor-camoufox-scraper... build
 
-COPY --chown=myuser package*.json ./
+# Self-contained production bundle; inject-workspace-packages copies the built @apify/scraper-tools
+# into node_modules instead of symlinking it. The native better-sqlite3 binding (a camoufox-js dep) is
+# compiled here against the base's Node 24, and the runtime stage uses that same base, so it matches.
+RUN pnpm --config.inject-workspace-packages=true --filter actor-camoufox-scraper deploy --prod /deploy
 
-RUN npm --quiet set progress=false \
-    && npm install --omit=dev \
-    && echo "Installed NPM packages:" \
-    && (npm list --omit=dev --all || true) \
-    && echo "Node.js version:" \
-    && node --version \
-    && echo "NPM version:" \
-    && npm --version \
-    && rm -r ~/.npm
+FROM apify/actor-node-playwright-camoufox:24-1.59.1@sha256:eadc96fa9492284eb45ef70b6b91c841fae7f142d25a22ae2887a21bb78b3469
 
-COPY --chown=myuser . ./
+# Base WORKDIR is /home/myuser and ships the Xvfb entrypoint plus the Camoufox binary in
+# ~/.cache/camoufox (outside node_modules). Replace only the template node_modules with the lean
+# production bundle (which carries the matching playwright + camoufox-js); the binary stays in the base.
+RUN rm -rf node_modules
+COPY --from=builder --chown=myuser /deploy ./
 
 ENV APIFY_DISABLE_OUTDATED_WARNING=1
 
-CMD ./start_xvfb_and_run_cmd.sh && npm run start:prod --silent
+CMD ["node", "dist/main.js"]

packages/actor-scraper/camoufox-scraper/package.json

@@ -10,9 +10,10 @@
         "@crawlee/playwright": "^3.14.1",
         "@crawlee/utils": "^3.14.1",
         "apify": "^3.2.6",
-        "camoufox-js": "^0.9.0",
+        "camoufox-js": "0.11.1",
         "idcac-playwright": "^0.2.0",
-        "playwright": "*"
+        "playwright": "1.59.1",
+        "playwright-core": "1.59.1"
     },
     "devDependencies": {
         "@apify/tsconfig": "^0.1.0",
@@ -22,19 +23,18 @@
     },
     "scripts": {
         "start": "pnpm start:dev",
-        "start:prod": "node dist/main.js",
         "start:dev": "tsx src/main.ts",
         "build": "tsc"
     },
     "repository": {
         "type": "git",
-        "url": "https://github.com/apify/apify-sdk-js"
+        "url": "https://github.com/apify/actor-scraper"
     },
     "author": {
         "name": "Apify Technologies",
         "email": "support@apify.com",
         "url": "https://apify.com"
     },
     "license": "Apache-2.0",
-    "homepage": "https://github.com/apify/apify-sdk-js"
+    "homepage": "https://github.com/apify/actor-scraper"
 }

packages/actor-scraper/camoufox-scraper/src/internals/crawler_setup.ts

@@ -18,7 +18,6 @@ import type { ApifyEnv } from 'apify';
 import { Actor } from 'apify';
 import { launchOptions } from 'camoufox-js';
 import { getInjectableScript } from 'idcac-playwright';
-import type { Response } from 'playwright';
 import { firefox } from 'playwright';
 
 import type { CrawlerSetupOptions, RequestMetadata } from '@apify/scraper-tools';
@@ -201,7 +200,11 @@ export class CrawlerSetup implements CrawlerSetupOptions {
                     ...this.input,
                     humanize: this.input.humanize ? Number(this.input.humanize) : 0,
                 }),
-            } as PlaywrightLaunchContext,
+                // `firefox` and @crawlee/playwright's PlaywrightLaunchContext resolve to two different
+                // playwright-core copies (camoufox pins playwright 1.59.1 to match its base image,
+                // while the rest of the workspace is on 1.61.0), so the BrowserType identities don't
+                // line up. This is a workspace dual-install, not a camoufox-js/crawlee issue.
+            } as unknown as PlaywrightLaunchContext,
             useSessionPool: true,
             persistCookiesPerSession: true,
             sessionPoolOptions: {
@@ -392,7 +395,7 @@ export class CrawlerSetup implements CrawlerSetupOptions {
 
     private async _handleResult(
         request: Request,
-        response?: Response,
+        response?: PlaywrightCrawlingContext['response'],
         pageFunctionResult?: Dictionary,
         isError?: boolean,
     ) {