Changelog

Plus 1 internal improvement for better reliability and performance.

Raise MAX_QUEUE_DEPTH default from 100 to 500 .

Add hard-cap eviction for _last_used_cache .

Standardize ResizeObserver height measurement .

Auto-dismiss activation banner when stats update after fi....

Resolve two TypeScript build errors .

Add lastModified to /smart-rendering-api sitemap entry (#....

Correct Phase H guard index in 0096 rollback .

Add herald.clarity_api_token to SOPS prod. .

Expand seo dedup guard to match slash-separated display n....

Update _CSR_WEBPACK_FILENAME_PATTERNS comment for dual-mo....

Show correct \$0.00 for zero-balance dormant users in ema....

Store create_task reference to prevent GC risk .

Replace containerHeight state with ResizeObserver ref in ....

Add SET NOT NULL to 0096 rollback Phase C .

Fall back to in-memory throttle on Redis failure in _shou....

Your scrape requests that hit the cache no longer cause 500 errors. Savings are now correctly set to 0, so billing works as expected.

Remove Site Crawler from homepage.

Plus 13 internal improvements for better reliability and performance.

Welcome modal not firing for ~13% of new signups .

Stop decrypting API key on every Mission Control load (#9....

Stop caching plaintext API keys in Web Storage .

Add rate limit and audit log to decrypt proxy .

Require explicit confirmation before API key transfer to ....

Stop storing plaintext API key in localStorage .

Mask API key in IntegrationCTA code snippets .

Update remaining /anti-bot-bypass-api references to /smar....

Align ToS and AUP with new marketing positioning .

Update stale slug-to-context mapping after renam....

Add 301 redirects for removed domain pages .

Hoist _POSTGRES_SURROGATE_TABLE to module level .

Correct llms.txt adaptive-rendering-api URL to smart-rend....

Replace bare console.log with debugLog to prevent job ID ....

Make UntranslatableCharacterError detection reachable via....

Expand _sanitize_for_postgres to cover Unicode non-charac....

Correct migration 0093 to target scrape_diagnostics for s....

Null out negative response_time_ms and scrape_time_ms val....

Replace time.time with time.monotonic for duration me....

Remove example URLs dropdown from config panel.

Add blog-post-headline/description classes to match speak....

Replace /anti-bot-bypass-api with /smart-rendering-api in....

Description: You can now run /usage to view your tier and linked email, and /balance to see your current dollar balance formatted as $X.XX. These commands are only visible to linked accounts.

Discord OAuth provider.

Discord linking UI.

Description...

Description...

Description...

Description...

Add blog-performance UI page + nav entry .

Harmonize BYOS data processor role and strengthen respons....

fix(api): add input validation for discord_id, provider allowlists (, ) – likely a fix, maybe impr.

fix(auth): guard against Discord users with unverified email – bug fix..

fix(auth): clear account_link_state cookie after linking completes – bug fix..

fix(auth): set account_link_state cookie as HttpOnly + Secure – security improvement? It's about .

feat(sentinel): add /usage and /balance slash commands for linked users – new feature..

feat(auth): add Discord to connected accounts linking UI – new feature..

feat(auth): add Discord OAuth provider to NextAuth – new feature..

fix(scraper): reject hostless URLs in batch schema and worker pre-flight – bug fix..

(Note: there are internal commits not needed.).

Input validation for discord_id and provider allowlists – improves reliability, maybe "Validate Disc.

Guard against unverified email causing 422 – improves auth flow..

Clear account_link_state cookie after linking – improves user experience..

Set cookie HttpOnly + Secure – security improvement..

Add /usage and /balance slash commands – new feature..

Add Discord to connected accounts linking UI – new feature..

Add Discord OAuth provider to NextAuth – new feature..

Reject hostless URLs in batch schema – improves scraper reliability..

New slash commands (/usage, /balance) – likely Features..

Add Discord to connected accounts linking UI – maybe Features or Improvements? It's a new UI feature.

Add Discord OAuth provider – maybe Features (or Improvements?). Probably Features..

Input validation for discord_id and provider allowlists – could be Improvements (enhancement) or Fix.

Guard against unverified email – fix? Could be "Fixed Discord email handling". Type:fix..

Clear account_link_state cookie after linking – fix? Could be "Reset link state cookie". Type:fix..

Set cookie HttpOnly + Secure – security improvement. Could be "Secure cookie flags for link state". .

Reject hostless URLs in batch schema – fix? Could be "Reject hostless URLs in batch requests". Type:.

Validate Discord ID and provider filters.

Handle unverified Discord email gracefully.

"Discord slash commands" = 3 words? Actually "Discord" (1) "slash" (2) "commands" (3) => 3 words, ok.

"Discord OAuth provider" = 3 words? "Discord" (1) "OAuth" (2) "provider" (3) => 3 words..

"Discord linking UI" = 3 words..

"Validate Discord ID and provider filters" = 5 words? Count: Validate(1) Discord(2) ID(3) and(4) pro.

"Handle unverified Discord email" = 4 words? Handle(1) unverified(2) Discord(3) email(4) => 4 words..

"Clear link state cookie" = 4 words..

"Reject hostless URLs in batch" = 4 words? Reject(1) hostless(2) URLs(3) in(4) batch(5) => 5 words, .

"Secure cookie flags for link state" = 6 words? Secure(1) cookie(2) flags(3) for(4) link(5) state(6).

Feature entries: icon:Rocket.

Improvement entries: icon:Zap.

Fix entries: icon:Bug.

Security entry: icon:Shield.

Slash commands: visualHint:screenshot.

OAuth provider: maybe code? But code hint is for code examples; not needed. Could omit..

Linking UI: screenshot..

Validation: stats? Maybe stats..

Email handling: maybe stats? Could use stats..

Clear cookie: maybe stats? Could omit..

Reject hostless URLs: maybe code? Could omit..

Secure cookie: maybe stats? Could omit..

Description...

Description...

Plus 7 internal improvements for better reliability and performance.

Clear link state cookie after linking.

Reject hostless URLs in batch requests.

Description...

Description...

Use correct AdminKPIRow cards prop and AdminStatCardProps....

Remove email from Discord lookup response and usage embed....

Add in-memory fallback for discord lookup rate limit when....

Make trust seed marker permanent and clear on invalidate ....

Defer /usage and /balance before resolve to avoid 3s ti....

Add URL host validation to ScrapePDFRequest and ScrapeOCR....

Make history empty state text mode-aware .

Migrate Gemini CLI from --allowed-tools to --approval-mod....

Prevent double-release of domain concurrency slot on inva....

Reject hostless URLs in batch schema and worker pre-fligh....

Prefer backend response_time_ms over client wall-clock in....

Remove leading space in min_confidence warning sentence (....

Allow input composition during send — keep submit guards ....

Secure cookie flags for link state.

Description...

Plus 13 internal improvements for better reliability and performance.

Resolve TypeScript narrowing error in blog chat publish b....

Remove leading space in min_confidence warning sentence (....

Allow input composition during send — keep submit guards ....

Disable publish/draft buttons while publish is in-flight ....

Rename segment_key → segment_slug in HeraldSegment interf....

Add isSending guard to prevent concurrent message sends (....

Hostless URL Rejection.

Users must now specify an approval mode when using the CLI, replacing the deprecated flag. This prevents accidental runs without proper policy approval.

Administrators can now generate blog drafts through an interactive chat interface directly in the admin panel. The system provides real-time topic suggestions and draft generation with visual loading states during processing.

The new publishing endpoints enable seamless transition from draft to published content with conversation history preservation. Each session maintains context for up to 24 hours, allowing for iterative refinement of posts.

Plus 17 internal improvements for better reliability and performance.

Admin panel usage statistics now display accurate credit usage data by using the correct column names in database queries.

Customer activity displays properly in the CRM pipeline by using the correct session tracking data instead of referencing non-existent columns.

Success rate percentages now display correctly in the admin panel by computing the proper percentage values instead of decimal fractions.

Billing settings and recharge information now show accurate data by querying the correct database columns for payment configurations.

Session integration success rates now calculate properly by computing the metric in SQL rather than attempting to select non-existent database columns.

A new admin page allows you to create, edit, and manage platform announcements that appear to users. The interface provides a centralized location for controlling what messages users see when they access the platform.

The announcement management tool is now accessible through a dedicated sidebar navigation item, making it easy to find and use the announcement features without searching through menus.

Monitor your AI workforce trust levels and performance with a new dashboard showing real-time trust gauges, task-specific performance metrics, and review queue management. You can now track autonomous system behavior and approve/reject AI-generated content through an intuitive Kanban interface.

Take full control of your Herald scheduling with a comprehensive monitoring dashboard featuring job status grids, success rate gauges, and manual trigger capabilities. You can view 7-day performance history, identify jobs with high failure rates, and run jobs on-demand from a single interface.

Manage your Reddit operations with dedicated overview pages showing KPI metrics, opportunity pipelines, post performance leaderboards, and subreddit fitness heatmaps. You can now track engagement metrics, monitor opportunity progress, and analyze subreddit performance from centralized dashboards.

Organize and manage your Herald user segments with new overview and detail pages. You can view member counts, average relationship scores, and last entry times for each segment, with detailed analytics for individual segments to optimize your targeting strategies.

Plan and manage your content workflow with a new calendar view offering monthly and weekly perspectives. The Kanban-style queue interface lets you approve or reject content in bulk with reason tracking, while inline previews help you make informed decisions about scheduled content.

Track your blog syndication performance across multiple platforms with a comprehensive dashboard showing platform health, cross-platform status matrices, and aggregate syndication counts. You can monitor publishing progress, identify failed syndications, and view detailed status for each published p

The system now tracks wasted time across scraping attempts and trips circuit breakers earlier when domains consistently fail. This prevents wasting resources on domains that are temporarily unreachable or blocking requests.

Per-domain concurrency limits prevent overwhelming domains with too many simultaneous requests. This results in more stable scraping sessions and reduces the likelihood of triggering anti-bot protections.

Error fallback pages now use shorter cache headers to prevent stale error states. The system also clears trip counts properly after manual resets, ensuring domains return to normal operation faster.

Mobile users now see properly sized footers that don't clip content, and the homepage announcement bar can be dismissed. These improvements create a smoother browsing experience on smaller screens.

Worker memory limits have been optimized to prevent out-of-memory crashes while maintaining performance. This results in more stable long-running scraping operations and better resource utilization.

Plus 22 internal improvements for better reliability and performance.

Experience more reliable admin panel interactions with improved error boundaries and toast notifications for API failures. Rate limit errors, authentication issues, and other API problems now provide clear user feedback instead of silent failures, making troubleshooting much easier.

Various stability improvements including proper cleanup of video event listeners, correct breadcrumb navigation, and fixed Open Graph image sanitization. These fixes address edge cases that could cause minor disruptions.

Critical stability improvements ensure admin panel features work reliably across all pages. Authentication mismatches, broken endpoints, and data display issues have been resolved, providing a consistent experience when managing your platform operations and user data.

Logo images have been optimized for faster loading, reducing file sizes from 290KB to 78KB while maintaining quality. This results in quicker page loads and better performance across all marketing and documentation pages.

The footer video now loads only when you scroll near it, preventing the 2.2MB video from buffering on initial page load. This significantly improves above-the-fold performance and reduces data usage for users who don't scroll to the bottom.

Marketing site navigation has been standardized across all pages using a shared navigation constant. This ensures consistent menu items and improves the overall user experience when navigating between pricing, documentation, and other sections.

The homepage now features a WebGL scroll-driven tunnel intro that creates an engaging visual experience as you scroll. This modern animation enhances the first impression while maintaining fast load times through optimized rendering.

Authentication-related pages now properly exclude the sticky footer video, providing cleaner layouts for password resets, email verification, and welcome pages. This ensures these transactional pages remain focused and distraction-free.

Logo sizes and spacing have been standardized across all headers and navigation elements. This creates a cohesive visual identity throughout the platform, with proper proportions between logos and wordmarks on every page.

Various minor issues have been resolved including backup retention validation, duplicate footer prevention on documentation pages, and improved scroll behavior for short marketing pages. These refinements enhance overall platform stability and usability.

Disposable email domains are now blocked at signup initiation using an expanded community-maintained list of 5,383 domains. This prevents abuse from temporary email services and improves platform security for all users.

The system now automatically detects when a domain is experiencing repeated failures and temporarily pauses requests to prevent batch job failures. This helps maintain overall scraping stability when encountering problematic domains.

Dashboard now shows your current tier with smooth count-up animations and branded icons that scale in when you scroll. The visual progression makes it easier to see your usage limits and available capacity at a glance.

A new brand resources page provides easy access to logo downloads, color palettes with hex and oklch values, typography specifications, and usage guidelines for partners and developers.

Comprehensive integration guides for Supabase are now available, including Edge Function and pg_cron examples, SDK walkthroughs for Python and Node.js, and common use cases like price monitoring and lead enrichment.

The billing interface now displays your current tier with the new branded tier badges and includes a progression hint showing your path to higher limits. This gives you clearer visibility into your usage and upgrade options.

High Volume tier users now see a clear upgrade path to Enterprise with custom limits tailored to their workload. The dashboard celebrates your tier upgrade with a special announcement when you have available balance.

All tier displays across the platform now use the same branded tier badge system with consistent colors and icons. This creates a unified experience whether you're viewing your dashboard, billing page, or tier progression.

Plus 51 internal improvements for better reliability and performance.

Password reset emails now include your display name when available, making communications feel more personal and professional. If no name is set, the email defaults to a friendly greeting.

Usage alert emails now show accurate dollar amounts instead of inflated values. The billing system correctly converts microcent balances to dollars, preventing confusion about remaining credits.

Display credits as dollar amounts in invoice upcoming ema....

Added dedicated startup scripts and entry points for automated processes, making it easier to deploy and manage scraping operations at scale. These improvements streamline the initialization of background services.

Implemented new configuration retrieval capabilities that allow seamless integration between different components. Users can now easily access and manage configuration settings across the platform.

Replaced the previous per-process rate limiting with a centralized system that ensures consistent limits across all instances. This prevents unexpected behavior during deployments and maintains fair usage policies.

Fixed the extension update mechanism to properly handle self-hosted installations. The update system now works consistently regardless of deployment environment, ensuring users always have the latest features.

Improved the connection detection logic to correctly identify when users are properly configured. This prevents false positives in the connection status and ensures the setup process works as expected.

Resolved font loading issues that affected visual elements in the platform. All necessary resources are now properly included in the deployment, ensuring consistent appearance across all features.

Enhanced the security of alert systems by properly sanitizing input data. This prevents potential security issues while maintaining the functionality of critical notification features.

Fixed configuration validation to use consistent constants throughout the system. This ensures that all components respect the same limits and prevents configuration drift between different parts of the platform.

API discovery now correctly bills for proxy usage when required, ensuring you're only charged for the resources actually consumed during the discovery process.

The system now correctly identifies captcha challenges even when JavaScript contains comparison operators, preventing false negatives in automated detection.

API discovery no longer attempts to use Next.js development-only paths, ensuring consistent behavior across all environments and preventing failed discovery attempts.

Webhook failures from DNS rebinding attacks now properly increment failure counts, triggering automatic disabling when thresholds are exceeded to protect your system.

The batch scraping interface now accurately reflects the 100 URL limit per batch call, preventing confusion and ensuring you can plan your scraping operations correctly.

Account deletion now properly revokes all session tokens before bulk deletion, ensuring complete account cleanup and preventing unauthorized access after deletion.

The system now correctly identifies rate-limited responses with short HTML bodies, ensuring appropriate escalation and preventing misclassification of blocked requests.

Deposits exceeding the maximum limit now trigger automatic refunds and block credit grants, preventing accidental overcharges and maintaining billing integrity.

Expired sessions are now correctly identified and filtered in the admin interface, providing accurate session management and preventing confusion about active sessions.

Webhook URLs are now validated immediately before delivery, preventing DNS rebinding attacks that could occur between creation and delivery time.

Stripe webhook events that arrive before database commits now trigger proper error responses, ensuring reliable credit granting and preventing data inconsistencies.

Various internal stability enhancements and performance optimizations have been implemented to provide a more reliable scraping experience.

Plus 17 internal improvements for better reliability and performance.

Geographic proxy targeting now supports over 30 additional country code top-level domains, providing better coverage for international scraping needs. This enables more precise location targeting for your proxy requests across a wider range of countries.

Legacy API keys with empty prefixes now show a meaningful 'sk_live_' prefix instead of appearing as 16 bullet dots. This makes it easier to identify and manage older credentials in the dashboard.

The API filter on the usage page now includes SDK and ecosystem sources like python-sdk, node-sdk, mcp-server, n8n-node, and browser-ext. This provides a complete view of all non-playground usage patterns in your account.

Amazon product pages with multiple sellers now extract price data directly from variant olp slot information, inline text, and script offer data. This includes new seller counts, lowest prices, and used item counts without requiring additional API calls.

The dashboard credentials panel now correctly displays the actual key prefix from the API response instead of constructing it from UUIDs. This ensures accurate representation of your API keys in the interface.

Backup processes now correctly handle nested command patterns in error messages, preventing incomplete or malformed alert notifications. This ensures you receive accurate and complete failure notifications when backup operations encounter issues.

Webhook retry mechanisms now properly manage loading states during recursive operations, preventing UI inconsistencies and ensuring reliable delivery of webhook notifications. This provides more predictable behavior when handling failed webhook deliveries.

Text extraction from web pages now includes additional safeguards against missing attributes, preventing errors when parsing complex HTML structures. This results in more reliable content extraction for your scraping operations.

API source filtering in the dashboard now correctly excludes crawl and batch jobs, providing more accurate job categorization and filtering options. This helps you better organize and view your scraping jobs by source type.

Image blocking settings now properly apply to additional domains including booking.com and instagram.com, giving you more control over bandwidth usage and page rendering for these specific use cases.

Legacy API keys now display with appropriate prefixes, preventing confusion about key types and ensuring you can easily identify your credentials. This improves the user experience when managing multiple API keys.

Backup operations to R2 storage now include performance optimizations and proper bucket handling, preventing upload failures due to permission issues. This ensures more reliable and efficient backup operations to your R2 storage.

Backup operations now include a hard storage limit to prevent unexpected billing overages on R2 storage. This provides better cost control while maintaining reliable backup functionality.

Browser pool configuration now supports legacy environment variable names, ensuring compatibility with existing setups while transitioning to new configuration standards. This prevents disruptions when updating your environment configuration.

Credentials are now handled more securely during backup operations by using environment variables instead of writing plaintext credentials to disk. This eliminates potential security vulnerabilities while maintaining the same functionality for your backup workflows.

Added domain-based proxy bandwidth blocking for ads and trackers, including 65 new domains like analytics services, ad CDNs, and consent managers, reducing unnecessary data transfer and improving scraping efficiency.

Fixed domain blocking logic to properly handle URLs with user credentials and URL fragments, ensuring blocked domains like ad networks and trackers are correctly identified and filtered regardless of URL format.

Raised auto-recharge runtime cap from $1,000 to $2,000 and deposit cap to $2,000 to match frontend pack offerings, eliminating validation errors for users setting higher billing thresholds.

Increased cache TTL to 60 minutes and optimized browser configuration loading, reducing redundant network calls and improving overall API response times for repeated requests.

Launched a redesigned /best-web-scraping-api landing page featuring a 45° ASCII benchmark table, canvas ASCII background, and visual ranked cards with animated progress indicators for better user engagement.

Introduced a new 3D rolling cubes animation component using pure CSS 3D transforms and requestAnimationFrame, replacing the previous CardStream component in the BenchmarkHero section for a more dynamic visual experience.

Plus 6 internal improvements for better reliability and performance.

The system now automatically retries fetchBatch and fetchWebhookLog operations when encountering stale API keys, transparently handling key rotation without requiring manual intervention from users.

Fixed key prefix display in CredentialsPanel and improved LiveTerminal functionality by properly fetching decrypted key values, ensuring users can accurately manage and use their API credentials.

Enhanced scrape route security by prioritizing Cloudflare's cf-connecting-ip header over client-supplied x-forwarded-for, making it harder for malicious actors to spoof their origin in proxy requests.

Plus 29 internal improvements for better reliability and performance.

Batch proxy routes now accept X-Key-Id as an alternative to X-API-Key, allowing the server to resolve plaintext keys on demand instead of relying on sessionStorage caching.

Added on-demand API key decryption for batch operations, preventing fallback to truncated UUIDs when keys aren't cached in sessionStorage.

Corrected batch form submission to use 'urls' field instead of 'requests', ensuring proper API request formatting for all dashboard batch submissions.

Worker job queue processing is now more responsive with a 10-second timeout limit, preventing indefinite blocking and ensuring tasks complete in a timely manner. This improvement reduces delays in your scraping workflows and improves overall system throughput.

The system now actively monitors idle database connections, alerting you when connection counts exceed safe thresholds. This proactive monitoring helps prevent performance degradation and ensures your scraping operations maintain optimal database performance.

Critical system components now handle missing environment configurations gracefully, preventing unexpected failures during deployment or configuration changes. This improvement ensures your scraping infrastructure remains operational even when certain optional services are temporarily unavailable.

System health checks now provide more accurate status reporting by properly handling missing service configurations. This improvement gives you clearer visibility into system health and helps you quickly identify and resolve infrastructure issues.

The system now uses memory more efficiently during high-load scenarios, preventing performance degradation during intensive scraping operations. This optimization ensures consistent performance even when processing large volumes of data or handling multiple concurrent requests.

Various minor issues have been resolved to improve overall system reliability and performance. These fixes address edge cases and improve the robustness of core scraping functionality.

Image processing now correctly handles various file formats and color modes during uploads. Problematic image modes are automatically converted before saving, preventing cryptic errors and ensuring consistent behavior across JPEG, PNG, and WEBP uploads.

Browser pool operations now properly synchronize access to shared resources, preventing potential conflicts when multiple processes interact with browser instances. This ensures more stable and predictable behavior during concurrent scraping operations.

Admin credit adjustments now use advisory locks to prevent race conditions during balance updates. This ensures accurate credit tracking and prevents potential inconsistencies when multiple administrators modify account balances simultaneously.

Rate limiting for admin operations now properly handles Redis connection failures by failing closed rather than allowing unlimited access. This maintains security boundaries even during infrastructure issues.

Audit log failures during privilege changes now generate proper error notifications instead of silent warnings. This ensures critical security events are properly tracked and administrators are alerted to potential issues.

Soft-deleted and fraud-flagged admin accounts are now properly blocked from accessing sensitive features including blog management, proxy services, and API endpoints. This closes potential security gaps where disabled accounts could still perform privileged actions.

Administrators can now adjust user credits up to 10 times per minute, preventing abuse of the credit adjustment system. This rate limit is enforced per admin user and provides better control over credit management operations.

All admin user modifications including credit adjustments, admin status changes, and fraud level updates are now recorded in an audit log. This provides complete visibility into who made changes and when, improving accountability.

Analytics widgets now load lazily using Intersection Observer, deferring data fetching until widgets scroll into view. This improves initial page load times for GeoDistribution, ErrorBreakdown, HourlyPatterns, and TopEndpoints widgets.

The best-web-scraping-api page now features smooth scroll animations and improved spacing. Section labels include primary-colored numbers and line separators for better visual hierarchy and readability.

Plus 16 internal improvements for better reliability and performance.

Log data and LLM outputs are now properly sanitized before processing to prevent prompt injection attacks. This includes wrapping domain health data in XML tags and sanitizing raw log lines, making monitoring more secure.

The welcome email and verification screen now show the actual credit amount awarded instead of always displaying $1.00. Users receiving reduced bonuses will see the correct amount in both surfaces.

Verify code responses now return microcents instead of millicents for credits_awarded, making it consistent with all other balance fields in the API. This ensures accurate credit tracking across the platform.

The welcome email now uses the correct API keys route and removes false fee warnings. The message focuses on getting started quickly with a clear value proposition instead of outdated pricing information.

The cookie consent banner no longer intercepts sidebar navigation clicks on dashboard routes. Users can now access Billing, Settings, and Herald sections without banner interference in fresh sessions.

Entering $0 in the custom deposit amount now shows the minimum deposit error message instead of silently disabling the Add button. Users receive clear feedback about the $10 minimum requirement.

Users with only signup bonus credits no longer receive low-balance alerts. The system now checks for paid deposits before sending balance warnings, preventing unnecessary notifications for free-credit-only accounts.

API key creation now enforces the 100-character limit at the schema level, preventing PostgreSQL string-too-long errors. Users receive immediate feedback instead of HTTP 500 errors when creating long key names.

Credit balance calculations now happen directly in SQL to avoid precision loss with large numbers. This ensures accurate balance tracking even for high-volume users and prevents rounding errors in financial calculations.

Administrators can no longer modify their own admin status through the user management interface. This prevents accidental self-lockout and potential privilege escalation scenarios.

Webhook delivery now uses isolated database sessions per delivery attempt, preventing connection issues during retry operations. This improves the reliability of webhook notifications for your integrations.

Broadcasts scheduled for later delivery now properly trigger the send operation. The system correctly handles both immediate and scheduled broadcast delivery, ensuring your messages reach their intended recipients.

The fraud suspect query now uses last seen timestamps instead of random UUID ordering, ensuring high-risk users are properly surfaced. This provides more accurate fraud detection and better prioritization of suspicious accounts.

User detail pages now correctly display fraud response levels and credit adjustment reasons. The interface properly handles all user fields and provides accurate information for admin decision-making.

Broadcast resend operations now correctly target the intended audience without requiring manual recipient specification. The system automatically handles audience targeting based on subscription status.

Admin authentication now includes audience validation and shorter token expiration (15 minutes instead of 1 hour). This provides stronger security for admin operations and reduces the window for potential token misuse.

A comprehensive job history page is now available at /dashboard/history, allowing users to browse all past jobs from PostgreSQL with advanced search, filtering by URL/status/mode, pagination, and response viewing capabilities. The sidebar navigation has been updated to include this new History entry

Internal refinements to the dashboard layout and component structure for improved consistency and maintainability.

Job status counting operations now use batch Redis MGET commands instead of sequential GET calls, reducing Redis round-trips from O(n) to O(1) per reconciliation. This optimization cuts total GET operations for a crawl from O(n²) to O(n), significantly improving performance for large-scale scraping

Rate limit checks now occur before expensive network operations during URL discovery, preventing over-limit API keys from triggering unnecessary sitemap or start-page requests. This change reduces wasted resources and ensures users receive immediate 429 responses when they exceed their rate limits.

API key scope fallbacks now include billing permissions by default, ensuring consistent access control across all API key operations. This change improves security and prevents unexpected permission issues when working with billing-related API endpoints.

Plus 5 internal improvements for better reliability and performance.

Low balance alerts and auto-recharge triggers are now properly wired into both inline/sync and async queued billing paths. The system now correctly handles balance checks and sends timely notifications when account balances run low, preventing unexpected service interruptions during active scraping

Spending by Type charts now show accurate costs instead of estimated values based on request counts. The Recent Activity section maintains consistent height even when empty, and fallback messages display correctly when per-type spend data is unavailable, providing users with reliable billing informa

The system now properly handles structured error metadata from workers, accepting dictionary error formats in CrawlPageResult instead of raising validation errors. This enhancement provides more detailed error information for debugging and improves the reliability of error reporting in the dashboard

Deposit pack cards have been consolidated into a unified flat-row layout, removing the previous split between Regular and Smart deposit sections. This streamlined design provides a cleaner, more consistent user experience when managing account deposits and viewing available payment options.

Amazon Australia timeout settings have been aligned with other Amazon locales, using consistent max_wait_ms and nav_timeout_ms values across all regional endpoints. This standardization ensures reliable scraping performance and prevents timeout-related issues when working with Australian Amazon site

Set a balance threshold and AlterLab automatically tops up your account when you dip below it. Choose your trigger amount, top-up pack, and preferred payment method — it just works in the background. Configure it from the new billing page.

Deposit packs now come with graduated volume bonuses — the more you deposit, the more free credits you receive. Tier badges show your current tier and exactly how much extra you unlock at each level.

The billing page has been rebuilt as a single-page layout with a new tier rail, combined transaction and events history, receipt tracking, and a live RecentActivity feed — all without needing to switch tabs.

Get an email notification before your balance hits zero. Alerts fire at configurable thresholds so you can top up before any scraping jobs are affected.

The platform now automatically promotes your oldest payment method as default when no default is set, and keeps auto-recharge synced when you change or remove payment methods.

All API keys now include the billing scope by default. Existing keys have been backfilled. This unlocks balance and auto-recharge endpoints for programmatic billing management.

Auto-recharge enable, disable, and delete actions are now recorded in the billing audit log. Recharge attempts track the full lifecycle — pending, success, decline, and idempotent skips.

Volume bonus credits granted at deposit time now appear as a separate line in your transaction history and on the deposit success page, so you can always see exactly what you received.

Plus 30+ additional fixes across billing, accessibility, rate limiting, and dashboard reliability.

Scraping timeouts for amazon.com.au were lower than other Amazon locale configs, causing more frequent failures on large product pages. Now aligned with amazon.com, .co.uk, and .de.

Fixed a suite of edge cases: atomic single-UPDATE for backoff state, NULL sentinel for pending recharge rows, partial unique index on payment intent IDs to prevent duplicate audit rows, and proper asyncio.to_thread wrapping for Stripe calls.

The Penetrator anti-bot tier's learned cookie storage now filters out authentication and session cookies. This prevents accidental credential leakage across domains in multi-tenant environments.

Reduced step 1 to step 2 drop-off by preventing accidental exits through disabled click-outside and Escape key dismissal. Moved the skip button to the bottom with de-emphasized text and added value proposition subtext to guide users through the process.

Plus 27 internal improvements for better reliability and performance.

Dashboard polling interval reduced from 5 seconds to 15 seconds and now pauses when browser tabs are hidden. This reduces unnecessary network requests while maintaining real-time updates when actively viewing the dashboard.

Fixed soft 404 error flagged by Google Search Console on the blog index page. The issue was caused by empty heading states when no posts were found, which has been corrected to properly display blog content.

Credit history now displays amounts in dollars instead of raw microcents, making transaction values easier to understand. Grant-type transactions are also included in credit history, and pagination has been added to recent activity for better navigation.

Broadcast promo notifications no longer expose promo codes in notification bodies, preventing unintended code sharing. User-targeted promos still include codes only visible to the intended recipient, with expiration times properly synchronized.

All audit action values now have corresponding labels and icons defined, ensuring consistent display throughout the platform. This resolves missing visual indicators in audit logs and activity tracking.

Implemented edge-level protection that blocks unauthorized POST requests to demo endpoints and rate limits demo traffic to 5 requests per 10 seconds per IP. This eliminates millions of bot requests daily that previously overwhelmed system resources.

Your account's login history now shows the city and country for each sign-in, letting you spot unfamiliar access at a glance. Location data is geo-resolved from the login IP and appears in the Login History tab under Settings.

This release includes 2 internal improvements for better reliability and performance.

Enhanced automated handling of interactive challenges including modern verification systems. The platform now automatically processes complex verification steps and caches authentication tokens for improved success rates.

Revised endpoint documentation to accurately reflect current scraping tier capabilities. Users now have clear guidance on demo limitations and upgraded features available in production tiers.

Plus 7 internal improvements for better reliability and performance.

Added global rate limiting to prevent distributed abuse of demo endpoints. The system now tracks total demo requests across all IPs, ensuring fair access while maintaining protection against coordinated scraping attempts.

A new automated system now evaluates pending review items against brand guidelines using advanced analysis. This ensures consistent quality standards across all submissions while providing clear reasoning for approval or rejection decisions, streamlining the review workflow.

Extraction accuracy for sofiaman.ro has been significantly improved with theme-specific selectors that match the actual DOM structure. The updated selectors replace generic targeting with precise element identification, resulting in more reliable data capture from Shopify-powered stores.

The demo environment now uses the same components as the main dashboard, providing a consistent experience across all interfaces. This unification eliminates discrepancies between demo and production environments, making it easier to test and validate scraping workflows.

Plus 8 internal improvements for better reliability and performance.

New admin dashboard provides detailed ROI tracking with per-channel metrics including ROI percentage, cost-per-acquisition, and signups per hour. View comprehensive activity ledgers with create, edit, and delete capabilities through an intuitive modal interface.

Capture and analyze attribution data from UTM parameters, referrers, and landing pages on first visit. Track user acquisition sources through enhanced signup flows and OAuth integration, with data stored in first-party cookies and session storage.

Python and Node SDKs now automatically include ecosystem identification headers in API requests. Track usage patterns across different integration methods including Python SDK, Node SDK, MCP server, n8n node, and browser extensions.

Access comprehensive analytics through new admin endpoints covering channels, cohorts, funnels, and ecosystem attribution. Query materialized views for channel performance, cohort retention, and funnel conversion metrics with real-time data refresh capabilities.

Receive automated weekly insights reports aggregating key metrics from channel performance, cohort retention, funnel conversion, and ecosystem attribution. Personalized reports delivered through the existing notification system with frequency controls.

New admin dashboard provides detailed ROI calculations, cost-per-acquisition metrics, and activity ledger with full CRUD operations. Track total costs, revenue, signups, and effort across all channels with date range filtering.

Comprehensive admin analytics API with endpoints for channels, cohorts, funnels, ecosystem attribution, and overview summaries. Includes materialized view support for fast query performance and view refresh capabilities.

SDKs now automatically send ecosystem source headers identifying integration tools like Python SDK, Node SDK, MCP server, and n8n. Attribution data is captured during signup and OAuth flows for better understanding of user acquisition channels.

Automated weekly task that queries all Pathfinder analytics views and delivers structured insights reports via Herald. Get regular updates on channel performance, cohort retention, funnel conversion, and ecosystem attribution without manual queries.

New event service provides robust behavioral tracking with Redis batching and background flush operations. Capture key user actions including signups, credit topups, subscription cancellations, and API key creations without blocking primary operations.

Modal dialogs now feature proper accessibility support including role attributes, focus management, and keyboard navigation. Enhanced user experience for cost management and activity tracking interfaces.

Plus 27 internal improvements for better reliability and performance.

Enhanced content scoping now properly handles image-heavy pages and domain-specific CSS selectors. Fixes issues with Readability stripping CSS classes and ensures promotional pages with images are correctly identified as valuable content.

Scrape responses now include optional warning and content quality fields, providing users with additional context about data reliability and potential issues during extraction.

Plus 9 internal improvements for better reliability and performance.

Platform balance information is now included in the Revenue section of monitoring reports, giving you clearer visibility into financial metrics and platform health.

IP geolocation services now use HTTPS by default through a more secure provider, ensuring all location data collection meets modern security standards and protects user privacy.

Banner classification logic has been adjusted to prevent misclassification of promotional images as hero content. This change improves accuracy for e-commerce sites with complex layout patterns.

The crawl Run button will no longer remain disabled after consecutive API failures. The system now tracks consecutive poll failures and automatically resets after 3 failed attempts, ensuring you can retry crawls without manual intervention.

The Python and Node SDKs now validate section_filter parameters on the client side, catching typos and type errors before they reach the API. This prevents common configuration mistakes and provides immediate feedback during development.

Extracted article text now has less boilerplate content like navigation, headers, footers, and aside elements removed. The improved cleaning process provides more focused content by stripping residual elements that Readability's heuristics might miss.

Image URLs containing commas are now parsed correctly from srcset attributes. The improved parser preserves commas embedded in URLs for services like Cloudinary and Shopify, preventing broken image links in extracted content.

You can now remove specific elements like navigation from scoped content using the new remove_selectors option. This is particularly useful for homepage scraping where the main content wrapper includes unwanted navigation elements.

Content containers rejected by the classifier now retain their valid child elements like paragraphs, lists, and headings. This prevents the loss of legitimate content that happens to be inside a container the classifier deemed non-content.

The system now resolves image sources from common lazy-loading patterns including data-lazy-src, data-original, data-image, srcset, data-srcset, and picture/source elements. This ensures images from Shopify themes and other lazy-loaded content are properly extracted.

Hero classification now recognizes carousel, slider, swiper, and slideshow elements in addition to traditional hero, banner, jumbotron, and masthead classes. This improves detection of main banner sections on e-commerce sites and modern web designs.

The quickstart hub page is now accessible, resolving the previous 404 error. This provides a proper entry point for new users getting started with the platform.

The Python and Node SDKs now support scroll_to_load and scroll_count parameters for handling infinite scroll content. This enables more comprehensive scraping of pages that load additional content as you scroll.

Log-monitor Discord reports now display all sections consistently, including Top Domains, API Activity, Recent Jobs, Active Users, Security, and Issues. No sections are skipped based on conditions.

Top-up notifications in Discord now filter by credit type instead of amount thresholds, preventing false alerts from free trial grants and refunds while accurately tracking genuine top-ups.

Discord reports now display users_new_1h when there are new signups in the past hour, providing real-time visibility into platform growth and user acquisition trends.

Platform failure rate metrics now exclude user-gated failures from tier budget caps, providing more accurate all-time failure statistics that reflect actual platform reliability rather than user limitations.

The scrape response now includes the missing note field, ensuring consistency with the unified response format. This alignment prevents data gaps when processing results through the Node SDK.

The cookie banner now appears as a floating card instead of sticky, creating a cleaner user experience. The sticky CTA is hidden until the cookie banner is dismissed, preventing interface conflicts.

Security improvements block IPv4-mapped IPv6 addresses and bracketed IPv6 formats that could bypass SSRF protection in demo routes. This closes potential attack vectors while maintaining legitimate functionality.

The SDK now validates section filter parameters before API requests, catching typos and type errors in min_content_blocks, exclude_content_types, and content_only options. This prevents failed requests due to configuration mistakes and provides immediate feedback during development.

Plus 3 internal improvements for better reliability and performance.

Carousel detection on Shopify sites now uses a more reliable two-pass strategy. The system first looks for data-section-type attributes, which Shopify themes use to annotate major sections like slideshows, providing higher confidence in identifying the primary carousel.

Hero image classification now correctly identifies only the primary carousel container instead of tagging every image in all carousels. This prevents excessive hero tags on pages with multiple carousels and improves image categorization accuracy.

Text extraction now automatically removes navigation, header, footer, and aside elements after Readability processing. This eliminates residual boilerplate content that previously appeared in extracted text, delivering cleaner article content and more accurate text analysis.

The playground interface now properly resets after API failures, preventing the Run button from getting stuck in a disabled state. The system tracks consecutive poll failures and automatically bails after 3 attempts, ensuring the interface remains responsive for subsequent operations.

The playground now handles cases where keys created in prior sessions or seeded keys don't have cached plaintext, preventing silent failures during crawl operations.

The Run button in the playground no longer remains disabled after consecutive API failures, as the system now properly tracks and resets polling states after multiple errors.

Login and signup patterns no longer strip legitimate standalone content from paywalled sites, preserving important text that was previously removed during extraction.

The scroll_count parameter is now only included when scroll_to_load is enabled, reducing unnecessary data in API responses and improving request efficiency.

SDKs now validate section_filter parameters against allowed keys and value types, preventing typos and invalid configurations from causing unexpected behavior during scraping operations.

Hero image detection now only applies to the primary carousel container, preventing every image in nested carousels from being incorrectly tagged as hero content.

Navigation, header, footer, and aside elements are now stripped from extracted text after Readability cleaning, resulting in cleaner content without residual boilerplate elements.

Commas inside URLs in srcset attributes are now preserved correctly, handling Cloudinary transforms and Shopify query parameters that were previously split incorrectly.

Scroll_to_load functionality now works seamlessly with render_js set to auto, providing more flexibility for scraping sites with dynamic content loading.

Image source resolution now handles common lazy-loading patterns including data attributes and srcset, improving image extraction from sites using various lazy-loading techniques.

Python and Node SDKs now support the section_filter parameter, keeping SDK functionality in sync with the API's SectionFilter model for more precise content extraction.

Add content depth events for docs and blog .

Wire product engagement events in dashboard .

Content extraction now supports removing navigation elements from scoped content using the new remove_selectors field. This provides cleaner results when scraping homepage content that includes navigation menus within the main content container.

Python and Node SDKs now support scroll_to_load and scroll_count parameters for JavaScript scraping. This enables automated scrolling for infinite scroll pages and provides better control over dynamic content loading.

Plus 1 internal improvement for better reliability and performance.

Defer content_only heading-strip until after child promot....

Throw on missing DEMO_TOKEN_SECRET in production .

Include scroll params in cache key .

Use callback refs in useScrollDepth/useTimeOnPage .

Gate FLAG_ICON_PATTERNS on small dimensions .

Add classifier-less fallback for json_v2 section tree (#3....

Level-aware pruning guard — only preserve h1/h2 headings ....

Unwrap options wrapper in UnifiedScrapeRequest validator ....

Fallback to original soup when Readability empties sectio....

Scope markdown with playbook content_selectors .

Add leaf-container fallback for json_v2 section tree .

Prune empty leaf sections from json_v2 section tree .

JavaScript-based scraping now properly handles iframe inclusion, shadow DOM flattening, and custom actions parameters. Requests that previously failed with TypeError errors due to missing function arguments will now execute successfully.

Image source set parsing now correctly handles URLs containing commas, such as Cloudinary transformations and Shopify query parameters. This prevents broken image URLs when extracting responsive images from srcset attributes.

Content classification no longer skips valid content within containers rejected by the classifier. Text blocks, lists, and headings inside non-content containers are now properly included in the final extraction results.

Image extraction now resolves sources from common lazy-loading patterns including data-lazy-src, data-original, data-image, srcset, data-srcset, and picture/source elements. This ensures images from Shopify themes and other lazy-loading implementations are captured correctly.

Hero banner classification now recognizes carousel, slider, and slideshow containers in addition to traditional hero classes. This improves detection of main banner sections on e-commerce sites using modern slideshow implementations.

Security measures now block IPv4-mapped IPv6 addresses and bracketed IPv6 formats that could bypass SSRF protection. This closes potential bypass vectors in demo routes while maintaining legitimate access.

Use OIDC trusted publishing for Python and Node SDKs.

Replace postMessage wildcard origin with window.location.....

Echo correlationId in legacy cookie capture responses (#3....

Add domain pinning for BYOS cookies in inline fast path (....

Pass byos_session_domain to T4 escalation call .

Type session-proxy payload as JWTPayload instead of Recor....

Guard validate_session against zero-knowledge sessions (#....

Use windowed Redis counters for scrape bombing detection ....

Wire BYOS context pool into session invalidation chain (#....

Use HKDF-SHA256 for encryption key derivation .

Add TTL-based expiry to RSA public key cache .

Stream session validation response to prevent OOM .

Add correlation ID to legacy cookie capture protocol .

Remove dead cookie-parser.ts with duplicate validators (#....

Replace postMessage wildcard targetOrigin with window.loc....

Your workspace settings now include an Activity tab showing a full audit trail of team actions — invitations, role changes, and configuration updates. Stay informed about everything happening in your workspace.

View workspace usage statistics at a glance with the new stats card. You can also resend invitations, leave workspaces, and set custom avatar URLs directly from settings.

Share promo codes via URL — signed-out users are guided through signup with the code pre-filled, and the deposit page automatically applies it. No more copy-pasting promo codes.

New users who receive a reduced welcome bonus now see a clear modal explaining why, with an option to unlock the full bonus. No more confusion about varying signup credit amounts.

When a promo code is active, the deposit page now shows the discount clearly and pre-fills the code from URL parameters. The apply button no longer triggers errors from browser event objects.

Plus 20 internal improvements including homepage visual polish, SEO structured data enhancements, and blog reliability fixes.

Fixed several issues with workspace switching — balance now refreshes correctly, dialogs close properly before secondary operations, and tab state no longer resets unexpectedly when switching contexts.

API Keys and Playground tables no longer truncate columns on smaller screens. Content is fully visible without horizontal scrolling cutting off important data.

Added per-IP and per-email rate limiting to login, OAuth, and verification endpoints. Verification codes are now hashed before storage, and brute-force protection with progressive lockout is active on all auth routes.

API keys are now rotated on each OAuth token exchange instead of re-exposing the same key. Localhost redirect URIs are blocked in production, and re-verification emails are sent after email changes.

Revoking a session now properly invalidates the associated JWT token, preventing continued access with stale credentials.

New users now get a streamlined onboarding flow with one-touch API key creation and a guided playground tutorial. Get from signup to your first scrape faster than ever.

Receive helpful emails at key moments — onboarding tips after signup, integration guides if you're using the playground but haven't connected via API, and re-engagement nudges. All emails include one-click unsubscribe.

The dashboard now highlights features you haven't tried yet. See contextual cards for schedules, monitors, and API integration with code snippets ready to copy.

All platform emails now include HMAC-signed unsubscribe links and a dedicated unsubscribe page, ensuring full CAN-SPAM compliance.

The pricing page now includes a detailed comparison table showing how AlterLab stacks up against alternatives on features, pricing, and capabilities.

Ten documentation pages now include interactive screenshots showing real dashboard UI, making it easier to follow along with guides and tutorials.

The entire platform now uses consistent 'requests' terminology instead of 'credits', making pricing and usage clearer. All pages, components, and docs have been updated.

Standardized balance formatting, latency display, and error states across the entire dashboard. Stat cards use smart formatting, and billing shows consistent pricing everywhere.

Recent Activity now supports Load More pagination, the deposit success page has a countdown auto-redirect, and pricing displays are unified across all billing sections.

Plus 30 internal improvements including mobile sidebar fixes, SEO title rewrites, cookie consent positioning, webhook navigation, and blog cover image enhancements.

Fixed an issue where the usage page could get permanently stuck on a skeleton loading state. The page now loads reliably every time.

Fixed inaccurate competitor pricing claims across comparison and marketing pages. All pricing data now reflects current published rates.

The signin page now shows email/password form and Microsoft OAuth. Check-email rate limit raised to 10/hour with proper 429 error surfacing. OAuth errors redirect with toast notifications.

New /vs pages let you compare AlterLab against specific alternatives side-by-side, with feature matrices, pricing calculators, and migration guides. Find the right tool for your use case.

New step-by-step tutorials covering real-time monitoring dashboards, WebSocket updates, AI agent and MCP integration, multi-page crawling, price monitoring pipelines, and cloud storage export.

Dedicated migration guides help you switch from ScraperAPI, ScrapingBee, and Apify with code examples showing equivalent AlterLab API calls for each platform's features.

All interactive elements now meet the 44px minimum touch target size, making the platform easier to use on phones and tablets.

Python and Node.js section headers on the SDK documentation page are now clickable links, making it easier to share direct references to specific SDK sections.

Plus 12 internal improvements including CI permissions, accessibility attributes, blog CTA placement, and pricing page sticky navigation.

Demo rate limits now show a signup CTA instead of a generic error. Comparison pages include signup funnels, and OAuth errors redirect back to the originating page with a notification.

Fixed several dashboard elements that appeared interactive but did nothing when clicked. All clickable-looking elements now respond or have been restyled.

Fixed an issue where demo API routes could be accessed by unauthorized external requests. Demo endpoints now properly enforce access restrictions.

Scrape and crawl endpoints now enforce per-API-key rate limits. This protects your account from accidental overuse and ensures fair resource allocation across all users.

You can now choose between light, dark, or system theme in the dashboard. The new 3-way toggle automatically follows your OS preference when set to system.

The billing system has been fully migrated to pay-as-you-go. All references to legacy subscription plans have been removed, and the dashboard now shows balance-based account tiers throughout.

All transactional emails from AlterLab now include a reply-to address, so you can reply directly to any email you receive from us.

Fixed an issue in the playground where rapidly clicking the Run button could fire multiple simultaneous requests. The button is now debounced and navigating away aborts in-flight requests.

Fixed an issue where verification emails showed an empty name in the greeting. Emails now use a proper placeholder when no name is available.

The batch scraping interface has been completely redesigned with a cleaner layout, real-time SSE progress streaming, content preview tabs, and a full result viewer. You can now clone batches, re-run failed items, and import URLs from CSV/TSV/JSON files.

You can now save and load batch configuration templates for repeated scraping tasks. A new analytics dashboard shows success rates, cost trends, and domain breakdown across your batch history.

Export scrape results directly to your cloud storage. Configure S3, GCS, or Azure connections in the new Storage Integrations settings page, then attach them to batch and schedule jobs.

A new Monitoring page lets you track changes on web pages over time. Set up monitors to detect when page content changes and get notified of differences.

You can now pause, resume, or delete multiple schedules at once. A new analytics dashboard shows schedule health rankings and credit cost forecasts. Search, filter, sort, and toggle between list and grid views.

You can now manually trigger a scheduled scrape with the new 'Run Now' button, without waiting for the next cron interval.

You can now clone any existing schedule or batch configuration with one click, making it easy to create variations of your scraping setups.

Batch and schedule pages now show an itemized cost calculator so you can estimate costs before running jobs.

You can now delete your account from the Settings page. A new danger zone section provides the option with clear confirmation steps.

You can now rename your API keys for better organization and regenerate them if compromised, without deleting and recreating.

A new profile tab in Settings lets you update your name and other profile information directly from the dashboard.

A new terminal component in the dashboard lets you interact with the API using CLI-style commands for jobs, stats, analysis, transactions, and data export.

Webhooks now automatically retry with exponential backoff on delivery failure. A delivery log shows the status of each attempt.

Fixed an issue where the password change page was inaccessible due to missing API routes. Password management now works correctly.

Fixed an issue where the API key list endpoint returned full key values. Keys are now properly masked in list responses.

Fixed several display issues with balance and billing: thousand separators now appear on all balance displays, sub-cent precision is hidden when zero, and deposit validation works correctly on the client side.

Fixed an issue where the playground page had horizontal overflow on mobile devices, making the response panel inaccessible.

Fixed the contact form being hidden on mobile viewports. The form is now fully accessible on all screen sizes.

Fixed an issue where some OAuth users experienced a sign-in loop caused by a database table mismatch. OAuth login now works reliably.

Fixed spend counter drift that could cause inaccurate usage reporting and billing sync timeouts.

You can now crawl entire websites with the new POST /v1/crawl endpoint. Progressive depth crawling discovers and scrapes pages automatically, respecting robots.txt. Supports up to 100,000 pages and 50 levels of depth, with billing per page.

Both the Python and Node.js SDKs now include crawl methods. Start a crawl, check status, and retrieve results programmatically with the same familiar SDK interface.

The playground now supports crawl mode with a visual config panel, depth slider, and animated multi-page results. The homepage demo shows live page discovery in action.

Crawl costs appear in your usage breakdown and transaction history. The crawl status page shows per-page cost breakdowns, session totals, and automatic refunds for failed pages.

A new visual workflow editor lets you build scraping automations on a drag-and-drop React Flow canvas. Connect scrape, HTTP request, webhook, email, and download nodes to create complex workflows.

A curated gallery of workflow templates helps you get started quickly. Browse pre-built templates and customize them for your use case.

Monitor your workflow executions in real-time with the run dashboard. View execution history, retry failed runs, and track progress.

You can now reset your password via email. The new forgot password page and backend flow lets you recover access to your account securely.

A new benchmarks page shows real performance data and comparisons, so you can see how AlterLab performs across different website types and protection levels.

Documentation pages are now served as markdown for AI agent consumption, along with a public OpenAPI spec and AI plugin manifest. This makes it easier for LLM-based tools to work with the AlterLab API.

Fixed an issue where usage logs paginated by raw database rows instead of top-level entries, causing confusing page counts when crawl jobs were present.

This release includes internal improvements for better reliability and performance.

The Crawl API is now live with POST/GET/DELETE endpoints, progressive depth crawling, and full playground support. Crawl results include per-page cost breakdowns and automatic refunds for failures.

Both Python and Node.js SDKs now include crawl methods. Start crawls, poll status, and retrieve results programmatically.

Crawl costs now appear in your usage breakdown and transaction history. The crawl status page shows tier breakdowns and session cost tracking.

You can now reset your password via a secure email flow. The forgot password link on the sign-in page walks you through the recovery process.

Export scrape results directly to S3, GCS, or Azure. Configure storage integrations in Settings and attach them to batch and schedule jobs.

You can now set spending limits per API key and at the account level. Scheduled and batch scrapes enforce these limits, and credits are automatically refunded when jobs fail.

Tier cards on the billing page are now interactive, making it easier to understand and compare what each tier offers.

Reduced initial page load time by deferring below-fold JavaScript and heavy 3D animations. Dashboard skeleton loaders replace spinners for smoother perceived performance.

Fixed an issue where some users experienced an infinite sign-in loop when using Google OAuth due to a database table mismatch.

The balance area in the dashboard header is now clickable and links directly to the billing page.

Crawl limits have been raised to 100,000 pages and 50 levels of depth, supporting larger site crawls.

Scrape as your logged-in self. Pass your own session cookies or custom auth headers to access member pricing, paywalled content, private communities, and internal dashboards. Zero credential storage — cookies are used for the request only and immediately discarded.

New 'cookies' parameter accepts cookie strings or key-value objects. Works with all tiers including JS rendering and anti-bot bypass. No surcharge for authenticated requests.

Pass Bearer tokens, OAuth headers, or any custom authentication headers via the 'extra_headers' parameter. Perfect for scraping authenticated API endpoints and web apps behind SSO.

Session data is held in memory only, encrypted via TLS 1.3, never written to disk or logs, and discarded immediately after the request completes. Your credentials stay yours.

Tier cards on the billing page are now interactive, making it easier to compare features and costs across scraping tiers.

Reduced initial page load by deferring below-fold JavaScript and heavy 3D animations. Dashboard pages now load noticeably faster.

Scraping is now more reliable with better handling of large HTML pages, faster dead proxy detection, and improved timeout management for browser-based scraping.

The balance area in the dashboard header is now a clickable link that takes you directly to the billing page.

You can now set daily and monthly spending limits on individual API keys and at the account level. Limits are enforced on scrape, batch, and scheduled jobs, with automatic credit refunds when jobs are blocked.

You can now apply promo codes when making deposits to receive bonus credits. Enter a code in the deposit form to see your bonus applied automatically.

You now receive email confirmations when deposits complete successfully, and reminders for abandoned checkouts.

Deposits now generate Stripe invoices. You can view promo bonus details as distinct line items in your payment history.

You can now specify a webhook_url per scrape request. Job completion payloads are enriched with more data for easier processing.

A new usage detail page lets you drill into request logs for each API key with pagination, filters, and CSV export.

A new domain analytics view in the dashboard shows success rates, costs, and trends broken down by target domain.

You can now pass an extraction_prompt parameter to have AI extract specific data from scraped pages. The LLM extraction feature processes your prompt against the page content and returns structured results.

The playground now includes an output format selector, defaulting to JSON and text. Choose which formats you want returned from each scrape.

A new RAG output format returns content in chunks with token counts, optimized for retrieval-augmented generation workflows.

A new monitoring layer on top of scheduled scraping detects content changes and diffs between runs, letting you track when pages update.

The dashboard, 3D visualizations, and terminal now fully support light mode with proper color schemes and contrast.

Export scrape results directly to S3, Google Cloud Storage, or Azure Blob Storage. Configure connections in the new Storage Integrations settings page.

A new LangChain community document loader integrates AlterLab scraping directly into your LangChain pipelines.

The Node.js SDK now includes batchScrape, waitForJob, getJobStatus, and estimateCost methods for programmatic batch operations.

Amazon scraping now extracts categories, BSR ranking, rating histograms, Amazon's Choice badges, delivery info, promotional offers, coupons, product variations, list prices, and savings data.

Scraping performance has been improved with cached extraction output, adaptive polling intervals, and connection reuse. Response times are noticeably faster, especially for repeated requests to the same domains.

Fixed multiple billing edge cases: duplicate credit grants on webhook retries, spend counter drift, incorrect low-credit threshold calculations, and proper microcent conversions across all billing paths.

Fixed an issue where the usage detail modal displayed incorrect tier, format, and location information for scrape requests.

The minimum deposit amount has been raised from $5 to $10 to ensure meaningful account balances.

Fixed scraping failures on sites with incomplete SSL certificate chains. These sites now scrape successfully with automatic certificate fallback.

When scraping fails, error responses now include clearer, structured details about what went wrong, without exposing internal system information.

You can now submit thousands of URLs in a single POST /v1/batch request. Track progress, get webhook notifications on completion, and view itemized billing for each URL.

Set up recurring scrape jobs using cron expressions. Schedules run automatically and apply your BYOP discount when configured.

New dashboard pages for managing batch jobs and schedules, with progress tracking, error details, and usage breakdowns.

A new DPA page is now available for enterprise customers who need a formal data processing agreement.

Fixed an issue where the usage chart on the dashboard was showing a placeholder instead of real data.

Failed scrape requests now appear in your detailed usage logs, giving you full visibility into what went wrong.

Internal performance optimizations to improve scraping reliability and reduce resource usage under high load.

Fixed an issue where signing in with Google OAuth could crash due to a logging error in the fingerprint service.

Updated extraction selectors for Amazon product pages to correctly capture coupon and promotional offer data.

New compliance layer screens signups against OFAC sanctions lists and applies geo-IP restrictions for prohibited regions.

The usage page now shows your current balance prominently and removes misleading hover effects from non-interactive elements.

Fixed several issues causing Google One Tap sign-in to fail, including session handling errors and redirect loops.

Fixed an issue where signing in with a different OAuth provider than originally used would show a confusing error instead of a helpful redirect.

You can now apply promo codes when depositing funds to receive bonus credits. Enter codes directly on the billing page during checkout.

A new notification bell in the dashboard alerts you to important events like promo code activations and account updates.

Webhook job.completed events now include richer data. You can also set a per-request webhook_url to receive results at different endpoints for each scrape.

Every deposit now generates a proper Stripe invoice, making it easier to track payments and manage accounting.

Better text and markdown output from structured data sources. Collection pages, financial data, and generic pages now return richer formatted content.

The AlterLab blog is now accessible from the main site navigation across all pages.

Scrape API responses are now faster thanks to adaptive polling and optimized backend processing. You should notice quicker turnaround on synchronous requests.

Fixed an issue where webhook retries from Stripe could result in duplicate credit grants to your account.

Fixed the usage detail modal showing incorrect format and location information for individual scrape requests.

Sites with incomplete SSL certificate chains no longer fail. The scraper now automatically falls back to handle partial certificate configurations.

Fixed the welcome modal displaying 1,000 free scrapes instead of the correct 5,000.

You can now enter promo codes when adding funds to your account. Valid codes grant bonus credits on top of your deposit amount.

A new notification center in the dashboard keeps you informed about promo activations, account events, and important updates.

Every deposit now generates a Stripe invoice with promo bonus shown as a distinct line item, making your payment history clearer.

Synchronous scrape requests now return faster thanks to adaptive polling optimizations and cached extraction output.

Sites with incomplete SSL certificate chains no longer fail. The scraper automatically handles partial certificate configurations.

Adjusted pricing for advanced scraping tiers to better reflect operational costs. Check the pricing page for current rates.

Fixed multiple issues where credit balances and costs were calculated incorrectly due to unit conversion errors. Your balance and usage now display accurately.

Fixed the usage detail modal showing wrong tier, formats, and location for individual scrape requests.

Product listings, search results, and other collection pages now return significantly richer structured data in both JSON and text/markdown formats.

Added optimized extraction for almaany.com, a major Arabic dictionary site, with dedicated content parsing rules.

When requesting text or markdown format, you now get richer output generated from structured data sources like product schemas and financial instruments.

Significantly improved success rates on Cloudflare-protected sites by reducing false challenge detections and eliminating wasted time on unsolvable challenges.