Work

The project updates its bundled Solana RPC websocket libraries and related UUID and validation packages to newer versions. This refreshes the dependency tree to improve compatibility and maintain support for current runtime requirements.

Updated the backend build and deployment pipeline to use the Debian Trixie Node.js 24 runtime base. CI and cache-layer generation now validate the OS codename to ensure only Trixie-compatible, CA-bundle-ready images are published.

The project’s GraphQL tooling packages were updated to newer versions. This refreshes the GraphQL schema merging and execution dependencies used by the application.

The project updates its styling library to styled-components 6.6.0. This refreshes related CSS-in-JS and Emotion peer dependencies to newer versions for more consistent compatibility.

The project updates its Socket.IO and Engine.IO client and server dependencies to newer patch versions. This improves realtime communication library compatibility while keeping the existing integration intact.

The project updated its dependency lockfile by removing several bundled Ethereum wallet and viem-related packages and their internal transitive entries. This narrows the resolved dependency graph for the build environment without introducing new code changes.

The repository updates core build tooling and related package versions to newer releases, including upgrading the TypeScript bundler and ORPC client dependencies. It also simplifies the TypeScript declaration build configuration to use automatic declaration generation.

1 commit · 7 filesAdded +196Deleted −151

The project updates the inquirer prompts package across backend and client dependencies. It also refreshes the lockfile to align related inquirer sub-dependencies for consistent installs.

The project updates its UI component framework dependency to a newer version and refreshes the associated lockfile entries. This aligns the app and UI package with the newer toolkit release.

The project updates its backend framework and related packages to newer compatible versions. It also refreshes the dependency lockfile to align transitive packages and tooling for the updated backend stack.

The API client documentation app was updated to render using a dedicated OpenAPI page component instead of a custom API client page. Dependencies were upgraded to newer versions to support the updated documentation rendering flow.

1 commit · 7 filesAdded +332Deleted −241

The project updates its MDX editor and related library dependencies, refreshing the lockfile to match. This enables integration with newer editor tooling for the admin interface.

The frontend environment configuration now loads through Next’s configuration entrypoint in isolated Node processes. This ensures the correct environment-specific settings, including expected origin registry hostnames, are available before bundling starts.

1 commit · 7 filesAdded +39Deleted −35

The project updates Ponder and related tooling to newer compatible versions and raises the required Node runtime. It also keeps required security fallbacks while adjusting deployment and cache migration guidance for the updated indexer.

The backend now verifies Privy tokens using a configurable local verification key with correct caching and preserves infrastructure failures. It also improves Privy API error mapping with consistent identity normalization and richer, sanitized response diagnostics.

1 commit · 31 filesAdded +1,297Deleted −1,171

The project updates its MDX remark plugins and dependencies to use native ESM, ensuring they can be loaded correctly in cold Node processes. Heading ID generation remains consistent for stable table of contents anchors.

1 commit · 9 filesAdded +46Deleted −49

The repository updates its CI action to run Turbo through Bun instead of Node tooling. The associated tests verify that changed workspaces are detected correctly across bootstrap modes without requiring an extra Bun setup in the caller.

1 commit · 3 filesAdded +42Deleted −11

Parked-domain and domain-forwarding behavior was updated to preserve the real request host and prevent search indexing on PR preview deployments. Metadata and redirect responses now consistently apply noindex, nofollow for these preview hosts.

1 commit · 7 filesAdded +122Deleted −10

The system now calculates domain expiration times with consistent precedence across ledger snapshots, captured email expiry, and registrar or NFT data. Tests were updated to verify correct selection when data is missing, conflicting, or stale.

1 commit · 4 filesAdded +115Deleted −28

The system now builds and runs the backend API, Temporal worker, and Next.js apps using Bun with a pinned Bun runtime. Deployment and preview environments were updated to start processes via Bun commands and reuse the same backend image.

1 commit · 51 filesAdded +210Deleted −259

The cart checkout story now waits for the two-year duration option to become visible and clickable before selecting it. This stabilizes the interaction and ensures the expected pricing amounts are asserted correctly.

1 commit · 1 fileAdded +7Deleted −2

A new admin bulk freeze feature lets staff freeze or unfreeze incident account and domain lists via pasted text or CSV while enforcing permission checks. The backend consistently blocks affected operations during review, queues required domain transfer locks, and exposes clear frozen-status messaging to users.

1 commit · 148 filesAdded +24,463Deleted −620

Frontend deployments can now optionally include a Datadog RUM application UUID. When provided, the build argument and resulting image tag include the PR and RUM UUID, ensuring the correct telemetry configuration is not reused across PRs.

1 commit · 4 filesAdded +94Deleted −1

API key create, revoke, rename, and restriction-update endpoints now emit structured audit records. Audits reliably mark success or failure and omit credential material such as signatures and plain API keys from logged content.

1 commit · 2 filesAdded +231Deleted −5

Logging now correlates emitted log records with active OpenTelemetry span context by including trace identifiers in both ordinary and audit entries when the span is valid. When no valid span is present, correlation fields are omitted to avoid misleading metadata.

1 commit · 2 filesAdded +76Deleted −0

Parked-domain rendering and proxying now resolve and preserve the correct request host, ensuring indexing decisions remain consistent. Preview deployments and domain-forwarding redirects are reliably served as non-indexable to prevent unintended crawler indexing.

1 commit · 7 filesAdded +122Deleted −10

The backend now validates API key IP and CIDR inputs more safely, bounding lengths and correctly handling IPv4-mapped IPv6 before parsing. Admin and workflow logic were adjusted to better respect paused states, avoid invalid statuses, and improve operational responses.

12 commits · 256 files

API key authorization now enforces strict client IP allowlists with validated syntax, separate IPv4 and IPv6 handling, and IPv4-mapped IPv6 normalization. The behavior is covered by new tests and applies whether or not an Origin header is present.

1 commit · 5 filesAdded +132Deleted −54

The project updates its Go x/net dependency to a newer version, reflecting corresponding changes in the module’s dependency checksums. This helps keep network-related libraries current for the application build.

Backend container builds now refresh OS packages during runtime dependency installation, then verify the CA certificate bundle. Tests were updated to ensure package-manager failures are not masked by cleanup and cached bases are handled safely.

1 commit · 3 filesAdded +50Deleted −39

Outbound HTTP transport now correctly manages connection pooling, request streaming, cancellation, and fetch replacement across Node and Bun. OpenTelemetry HTTP and Undici tracing are unified and preserve parent context without duplicating spans.

2 commits · 24 filesAdded +1,300Deleted −251

The frontend now loads server-resolved locales and delivers only feature-specific message namespaces to the client. This provides consistent server-rendered translations with isolated client scopes, preserving formatting and state across navigation.

1 commit · 52 filesAdded +808Deleted −188

Added an admin bulk freeze feature that freezes or unfreezes accounts and domains, with mixed-list CSV/paste input and result reporting. Implemented backend guards so frozen or under-review accounts and domains are blocked from protected API actions, with frontend read-only notices and retries for safe remediation.

1 commit · 137 files

Storybook smoke and test workflows now allow configuring Vitest parallelism via VITEST_MAX_WORKERS. Local tests default to two workers, while CI runs with four unless overridden.

Temporal refund and ledger transition runtime tests were updated to ensure workflow workers start consistently and that retry logic around Stripe creation and finalizer failures is exercised and replayed deterministically. This stabilizes replay expectations and gate behavior under bounded waits.

1 commit · 2 filesAdded +92Deleted −109

The update standardizes dialog and sheet content composition across the admin and frontend, improving consistency of spacing and structure. It also enhances record inspection layouts and refines interactive focus behavior through updated tests and Storybook stories.

1 commit · 336 filesAdded +15,493Deleted −9,754

The app delivers a streamlined domain search experience that preserves the input and focus while navigating between the homepage and results. It adds robust transfer mode handling, including bulk transfer and CSV import with authorization code parsing and validation.

17 commits · 67 filesAdded +4,199Deleted −974

Added Storybook coverage and validation for Arabic nested lists in article prose, ensuring correct right-to-left direction and logical padding. Updated prose styling to better support list indentation consistency in RTL contexts.

1 commit · 3 filesAdded +48Deleted −8

The order details payment receipt now omits approved refund labels and keeps net paid hidden when refund or evidence data is unavailable. Receipt totals and processing fees remain based only on recorded payment information and confirmed refund history.

1 commit · 19 filesAdded +113Deleted −231

The change hardens Storybook contract and autodocs tooling to handle missing base references more gracefully. It also adjusts related tooling behavior for git execution and environment-based base ref selection to reduce fragile CI failures.

1 commit · 7 filesAdded +21Deleted −112

The repository updates its tooling configuration to support TypeScript path aliases during runtime bundling. It also adjusts static analysis to ignore generated binary fixtures from certificate probing tests.

1 commit · 2 filesAdded +4Deleted −0

The frontend build now supports TypeScript path aliases during Node-side execution via JITI. This improves compatibility for modules that rely on tsconfigPaths.

1 commit · 1 fileAdded +1Deleted −0

Repository test execution was unified around Vitest, with consistent workspace test scripts, centralized tooling and offline infrastructure checks, and updated CI to run affected suites with controlled concurrency. Browser and E2E runs were adjusted to install Chromium/Playwright reliably and keep coverage aligned with the same test scopes.

1 commit · 161 filesAdded +2,152Deleted −1,463

The repository updates CSS build configuration to explicitly scan only runtime sources per app, excluding tests and story fixtures. It also hardens the frontend server tRPC client by enforcing required server-only auth key validation before making requests.

1 commit · 52 filesAdded +288Deleted −197

Admin users can now list saved order refunds and safely resume unfinished refunds after late or failed provider outcomes. The backend adds webhook-driven status refresh, verified repayment reconciliation, and new refund item models to ensure refunds are prepared, dispatched, and reported with correct amounts and permissions.

1 commit · 161 filesAdded +37,338Deleted −1,523

The frontend app now compiles Tailwind and shared UI styles once via the root stylesheet, with explicit source scanning for runtime code. Tests ensure the compiled CSS includes shared tokens and runtime files while excluding tests, stories, and showcases.

1 commit · 7 filesAdded +114Deleted −34

Cart and wishlist operations are now serialized through a shared queue with busy tracking and coordinated query invalidation. Errors are better handled with toasts and user-facing messaging, and optimistic updates are consistently reconciled or rolled back, including during account changes.

1 commit · 65 filesAdded +1,968Deleted −1,265

The frontend now evaluates operational feature flags via PostHog’s OpenFeature browser provider instead of LaunchDarkly, using a memory-only client that avoids analytics, replay, and consent waiting. Per-browser experiment overrides still take precedence while provider failures fall back to code defaults.

1 commit · 28 filesAdded +892Deleted −141

Session replay recording is now selected and started solely based on resolved measurement consent plus provider sampling and route/search triggers. The separate session replay feature flag and associated kill-switch behavior were removed, simplifying gating and test coverage.

1 commit · 9 filesAdded +43Deleted −184

Removed off-by-default experiment and feature flags for cart domain setup bulk copying. Bulk setup selection and apply-to-others behavior is now unconditional, with updated UI strings, Storybook, and tests.

1 commit · 28 filesAdded +190Deleted −227

The system adds a new daily “Drop Your Domains” workflow that prepares reviewed caption and matching GPT-generated poster images, caches generated assets immutably, and publishes them to X with stored media IDs. A new shared image-generation capability and dashboard endpoints support preparing, previewing, and scheduling this content.

1 commit · 57 filesAdded +10,152Deleted −362

The system now runs with updated AI SDK and Trigger dependencies, including an OpenTelemetry adapter registered for AI SDK telemetry. News draft/review execution was adjusted to preserve correct scheduling semantics, reuse compatible factual review versions, and validate analytics chat history with a contract-checked streaming response.

3 commits · 25 filesAdded +788Deleted −248

The software updated Available Gems and Domain Sales post generation to use consistent plain-writing rules and colon-based price lines. It also introduced a new Gems prompt version that lists bids without links or public URLs, while preserving validated internal sources.

1 commit · 10 filesAdded +87Deleted −27

The repository updates the X news collection process to treat missing or obsolete tweet lookups as non-fatal when possible. It also improves observation failure detection and warning reporting using the correct error set.

1 commit · 3 filesAdded +253Deleted −29

News discovery now attempts OpenAI Flex processing for scheduled GPT-6.1 Terra runs and retries capacity errors with exponential backoff. If Flex remains unavailable, it falls back to Standard once while preserving the gathered evidence and respecting the research abort signal.

2 commits · 13 filesAdded +506Deleted −35

News draft production now remembers the reviewed selection when the process hits the step or approval limit, and still saves assessments and keeps unfinished or unread leads available. It avoids carrying over reviewed selections from surplus submissions, later reads, or runs that fail after selection.

1 commit · 2 filesAdded +86Deleted −8

The agents now more accurately curate available gems using domain-only inputs and more reliably replenish changed or invalid queued evidence. News drafting was adjusted to better reuse coverage, track editor-scoped assessments, and handle pagination without skipping items.

1 commit · 9 filesAdded +593Deleted −97

Domain sales collection from DNScroll now validates individual table rows by source currency, preserves accurate USD rankings across mixed-currency pages, and avoids pagination issues at offset ties. It also performs a controlled Exa fallback after Cloudflare challenge or timeout and logs fallback reason and Exa cost details.

2 commits · 3 filesAdded +231Deleted −47

The system adds an end-to-end admin order refund workflow with saved approval records, verified repayment status checks, and safe resume/cancel behavior. It includes Stripe refund webhooks, accurate refund amount/tax calculations, and buyer/admin APIs that report what is confirmed versus under review.

1 commit · 158 files

The system introduces a robust order-refund model with new backend services, Temporal workflows, and admin UI for approving and resuming refunds based on audited saved approvals. It adds Stripe refund webhooks and reconciliation logic to reliably track repayment outcomes and keep user-facing refund messaging accurate.

1 commit · 158 filesAdded +36,714Deleted −1,197

Browser performance telemetry is now initialized with long-task tracking enabled. Static asset and API performance data are sanitized and redacted to restrict captured URLs and prevent sensitive header exposure while preserving useful timing and cache information.

1 commit · 7 filesAdded +244Deleted −14

The backend configuration now lowers the weekly free run limits for studio and lead generation tasks across canary, development, local, preview, production, and test environments.

1 commit · 6 filesAdded +30Deleted −30

Ledger transfers, DNS writes, and domain move flows now consistently enforce caller-specific holder proof and view-scoped organization access. Only the actual receiving holder can automatically accept based on destination control, preventing authorization bypasses from organization membership or display context.

1 commit · 109 filesAdded +4,848Deleted −3,164

Repository changes enforce consistent, holder-specific authority rules across organization and personal contexts for domain transfers, offers, and DNS writes. Auto-accept and cancellation behaviors now require caller-controlled destinations and accurate recipient identity permissions, with audit attribution preserved.

1 commit · 110 files

Added a dedicated data-synchronization operation to the virtual filesystem, ensuring it provides a data durability barrier without forcing a full flush. Updated the fd_datasync syscall to use this new mechanism, verified by tests that distinguish data sync from flushing.

1 commit · 3 files

Daily ingestion now decodes percent-escaped UTF-8 input, converts internationalized labels to Punycode, and validates then deduplicates encoded names without requiring literal counterparts. Atomic publishing and retry behavior are covered by new tests to ensure inventory remains unchanged on failed loads.

1 commit · 4 filesAdded +75Deleted −4

Domain settings and ledger transfer lifecycle actions are now audited with stable, connection-period attribution at admission. Audit streams include worker/Temporal events and avoid exposing rejected operations or uncertain workflow failures.

1 commit · 57 filesAdded +2,737Deleted −569

Domain search now preserves ranked TLD ordering while supporting additional unranked suffixes, validating and filtering them consistently across backend APIs and AI planning. Fast-path detection and Jev planning both respect the allowed TLD set so suggestions only include supported extensions.

1 commit · 19 filesAdded +227Deleted −81

Backend TLD pricing caching was updated so Route53 eligibility changes produce separate shared and in-memory results. The frontend was adjusted to improve pricing display consistency with loading states and formatting. Tests and eligibility logic were updated to cover the new behavior.

1 commit · 10 filesAdded +318Deleted −217

Search suggestions and AI-generated domain candidates are now constrained to registrar-supported suffixes fetched from a shared eligibility catalog. UI and WebMCP flows preload and respect this eligibility, preventing unsupported TLDs from appearing in results.

1 commit · 40 filesAdded +1,304Deleted −654

Implemented a cohesive search-to-transfer experience where the search input persists across homepage and results navigation. Added strict transfer parsing with authorization-code handling plus CSV/bulk transfer import dialogs that preserve drafts, with updated UI copy and animations.

16 commits · 68 filesAdded +4,209Deleted −971

The skip-auth banner was refactored into a persistent warning toast with an explicit exit action to disable the dev-mode auth bypass. Automated tests were added to verify correct toast behavior, cleanup, and demo-mode suppression.

1 commit · 2 filesAdded +172Deleted −38

Domain suggestions now respect the current registrar-supported TLD catalog rather than a static preference list. Both AI and manual search pathways filter, rank, and paginate generated candidate domains using the allowed registration-eligible suffixes.

2 commits · 36 files

Updated domain sales report rendering to show only the selected reporting source in the public heading. Adjusted the prompt version and tests accordingly so generated posts no longer include marketplace selection wording.

1 commit · 3 filesAdded +8Deleted −17

The domain sales collector now detects Cloudflare challenges from DNScroll and retrieves the challenged pages through Exa Contents using the configured API key. It parses Exa-rendered HTML without pagination and still allows correct ranking using a conservative coverage boundary.

2 commits · 3 filesAdded +300Deleted −27

The project updates its default news research, editing, and review model to GPT-6.1 Terra while preserving routing for focused corrections on GPT-6 Luna. It also improves usage reporting and adds tests for pricing and model selection correctness.

1 commit · 6 filesAdded +101Deleted −13

The ingestion pipeline now performs bounded, batched cleanup of obsolete snapshot artifacts while preserving protected receipts and retryable failures. Cleanup failures no longer block progress, and failed deletions are rotated for later retry without withdrawing accepted data.

1 commit · 12 filesAdded +125Deleted −1,031

The ingestion catalogue now refreshes using IANA’s delegated TLD list, filtering sources by final domain labels and preventing downloads for ineligible or inactive sources. If the IANA list is invalid or unreachable, the previous catalogue and accepted inventory are preserved until the next valid refresh.

1 commit · 4 filesAdded +72Deleted −4

The release adds a retirement migration that removes unused legacy archive tables and associated materialized data while preserving the current membership inventory and existing receipt and trend tracking. Tests and documentation are updated to reflect the new migration lifecycle and expected database state.

1 commit · 6 filesAdded +36Deleted −26

The project adds security-focused automation for dependency, secrets, and static scans with pinned tooling, while tightening DNS resolution to validate and properly encode inputs. Build and dependency baselines were updated to supported Node and refreshed packages to reduce known vulnerabilities.

1 commit · 17 filesAdded +1,026Deleted −1,187

Domain suggestion and AI search generation now filter candidates and UI TLD options by the currently supported registrar suffix catalog. Unsupported or ineligible extensions are excluded, and availability searches avoid broadening results when eligibility data is empty or fails.

1 commit · 34 files

Backend pricing and discovery reads now share only in-flight work while isolating returned rows and avoiding stale cross-request mutations. Redis cache reads/writes use bounded deadlines and safe fallbacks so cache connectivity issues no longer affect request correctness.

1 commit · 8 filesAdded +456Deleted −25

The Datadog frontend dashboard was updated with new paired comparison tables that relate connection baseline metrics to page loading, interaction, and API response timings using matched records. These tables display percentiles and matched sample counts, excluding data without a corresponding recent baseline.

1 commit · 3 filesAdded +513Deleted −3

The project updates its HPKE library dependencies to newer compatible versions in the lockfile. This aligns related packages and ensures consistent package resolution.

Updated the Datadog dashboard JSONs so group sections are no longer collapsed. Added a test to ensure dashboard group definitions only use properties supported by Datadog imports.

1 commit · 3 filesAdded +24Deleted −6

Unavailable search cards now show a secondary outlined import icon button for importable domains, hiding the transfer-style owner hint. Related stories and tests were updated to verify the new button and its accessible name across responsive card layouts.

1 commit · 18 filesAdded +53Deleted −56

The backend and frontend now emit an uncached performance ping with Server-Timing and trusted CORS controls. Browser instrumentation gathers timed baseline HTTP probes, separates server processing from request wait, and populates Datadog RUM contexts and dashboards with measured slow-experience cohorts.

1 commit · 18 filesAdded +4,067Deleted −1,178

Backend zonefiles lookups now use a dedicated timeout budget with stricter validation and cleaner connection pooling behavior. Cancellation is distinguished from timeouts so registrar checks and streamed availability results continue independently and consistently.

1 commit · 9 filesAdded +147Deleted −59

The deployment GitHub Actions job was updated to run on Ubuntu 24.04 instead of the previous Ubuntu image. This changes the runner environment used for deployment.

All CI/CD workflows were standardized on Ubuntu 24.04 and upgraded to newer GitHub Actions versions, including deployment tooling and common helper steps. This improves runner consistency and action/runtime compatibility across builds, tests, approvals, and deployments.

Domain search cards now present taken domains as unavailable with a small owner-only import link instead of a full registration/import pricing flow. Search availability logic and UI tests were updated to preserve taken registrar results across late DNS and pricing races.

2 commits · 21 filesAdded +211Deleted −115

The UI now uses a shared InfoTooltip component to render consistent informational help across the frontend. Tooltips are accessible via hover, focus, and activation, and the previous InfoPopover implementation was removed in favor of the unified behavior.

1 commit · 28 filesAdded +378Deleted −527

Zonefile lookups now record elapsed time and batch size with detailed stage-specific failure information when falling back to DNS. Timeout handling distinguishes connection versus query timeouts in warnings for easier debugging.

1 commit · 3 filesAdded +47Deleted −5

The search availability pipeline now optionally screens suggested domains using a read-only zonefiles snapshot database, emitting zonefile evidence alongside AI generation. DNS and registrar checks run independently, with client-side streaming and cache-aware DNS freshness.

1 commit · 25 filesAdded +1,177Deleted −157

Domain renewal flows now use domain availability information for renewal term validation and pricing. Reviewed renewal terms are enforced in the UI and checked against existing cart items, preventing silent price or year changes during add-to-cart and checkout.

1 commit · 16 filesAdded +429Deleted −202

Renewal flows now use per-domain quoted renewal pricing and compute correct annual and multi-year totals, keeping unavailable quotes unknown instead of substituting promotional estimates. The UI and tests were updated so partial availability blocks submission and shows accurate partial and resolved pricing.

1 commit · 18 filesAdded +1,376Deleted −152

Cart setup option switches for DNS-dependent features now remain disabled when custom nameservers are provided, with aria state reflecting “off” and selection blocked. Help tooltips for these features are still operable to explain the requirement.

1 commit · 3 filesAdded +155Deleted −18

Zone snapshot normalization now explicitly excludes only the known malformed tz entry while preserving total raw count validation. It also reports excluded entries, ensuring provider totals remain consistent without letting missing domains be hidden.

1 commit · 3 filesAdded +38Deleted −4

The loader now validates percent-encoded DNS names by requiring exact literal ASCII or Punycode counterparts, rejecting lossy or malformed encodings. Verified encoded duplicates are normalized and tracked, and failures preserve prior inventories without publishing incomplete updates.

1 commit · 4 filesAdded +84Deleted −3

The system now validates DNS and snapshot inputs more strictly, rejecting malformed or ambiguous names and handling permanently invalid sources without publishing empty inventories. It adds an independent, bounded health-check run and refines failure handling, freshness-aware lookups, and deployment timing for more reliable operation.

1 commit · 11 filesAdded +116Deleted −141

Ingestion now detects sources with missing provider domain lists and marks them unavailable without publishing empty inventories or failing other zones. The loader preserves previously accepted memberships and receipts, restores missing/corrupt artifacts from backup, and keeps lookups reporting unknown for unavailable sources.

3 commits · 8 filesAdded +76Deleted −18

The repository implements a new PostgreSQL-backed pipeline that ingests validated DNS zone snapshots, computes daily addition/removal aggregates, and serves discovery-style brand/search features. It adds automated ingestion supervision, health checks, and refreshed web views for namespace, TLD trends, and evidence-based availability.

2 commits · 25 filesAdded +2,941Deleted −1,790

The repository adds a new ingestion and publication pipeline that validates, receipts, and atomically publishes daily and snapshot changes with bounded retries. It also expands database schema and exposes a stricter /api/health and authenticated /api/lookup with freshness policies.

2 commits · 39 filesAdded +3,843Deleted −2,580

The repository adds automated diagnostics that fetch pinned published WASIX engine packages, run nested SQL exception probes, and validate expected results. It also verifies a retained Windows V8 engine prototype preserves the same SQL exception and execution behavior.

1 commit · 4 files

Tooltip components now render without animation while using consistent detail styling and sizing, removing the raised surface dependency. Stories and tests were updated to verify tooltip stability and correct open/close behavior.

1 commit · 6 filesAdded +31Deleted −9

The checkout UI now provides an accessible Holder ID explanation via an info popover next to the Receiving Holder ID field. The new help text is added for multiple languages and covered by an updated cart checkout story.

1 commit · 15 filesAdded +83Deleted −15

The frontend test suite and story assertions were updated to reflect UI changes. Loading screen expectations now check for “Receiving Holder ID” instead of “Receiving account,” and mocked components include the DomainCardSkeleton stub plus a privy runtime readiness flag.

1 commit · 3 filesAdded +3Deleted −1

The system now provides an organizations area with member, invitation, and activity management plus organization-scoped domain operations. Organization context is propagated across UI flows and authorization layers, including domain and DNS handling under organization permissions.

1 commit · 265 filesAdded +17,818Deleted −1,422

The app adds a full organizations workspace with member management and organization-specific domain controls. It also improves audit logs filtering and updates UI routing, data scoping, and translations to consistently support organizationId across domains, transfers, and DNS/DNSSEC workflows.

1 commit · 169 filesAdded +7,863Deleted −1,248

The backend now enforces organization- and connection-scoped permissions for domain reads, DNS writes, and ledger transition actions. It also records structured audit and activity data with organization and connection filters, and supports organization consent policies for holder connections.

1 commit · 46 filesAdded +4,176Deleted −408

The UI now consistently uses shared loading components, including accessible loading states and skeletons, across admin and frontend pages. Loading behavior was standardized to improve perceived performance and reduce inconsistent spinner usage.

1 commit · 373 filesAdded +13,532Deleted −5,230

The Listmonk container was updated to a pinned version and startup now reliably applies upgrades before starting the server. Static directory configuration handling was adjusted to use the updated environment variable name.

1 commit · 3 filesAdded +21Deleted −14

Introduces database-backed organization management with defined role and scope enums. Creates organizations, membership, and holder-identity tables with validation checks, relational constraints, and unique indexes to support controlled access and identity reconnection flows.

1 commit · 6 filesAdded +17,284Deleted −0

Order and notification list endpoints now preserve PostgreSQL timestamp precision in pagination cursors, including same-millisecond ties and recovery from legacy millisecond cursors. Cursor validation rejects malformed or mismatched date cursors without querying the database.

1 commit · 11 filesAdded +600Deleted −54

MPP and related order creation now ensures Temporal workflow startup only occurs after the database transaction commits and Temporal preflight checks pass. Workflow retries are idempotent and avoid restarting executions that already closed, improving recovery and preventing orphaned workflows.

1 commit · 10 filesAdded +566Deleted −224

Checkout now always validates selected cart items against live pricing rules, assigns the server-selected registrar, and uses exact database row versions to prevent changes during order placement. Order creation waits for the account lock, then consumes only matching rows before payment creation.

1 commit · 6 filesAdded +428Deleted −160

Updates Next.js response header handling and adds tests verifying redirects emit the correct Location header for cold, warm, and concurrent requests. It also confirms cookies, HTML, Flight responses, and not-found results preserve expected headers, including multiple Set-Cookie fields.

1 commit · 4 filesAdded +308Deleted −1

The system adds organization mode for managing domains across shared organizations using scoped roles and holder IDs. It introduces organization onboarding via invitations, server-paged UI with permissions gating, and backend enforcement for reads, transfers, exports, DNS updates, and auto-renew behavior.

1 commit · 75 files

Added a full organizations experience with member and activity management plus navigation and SEO controls, gated by admin permissions. Updated domain, DNS, and ledger transition flows to respect organization-scoped Holder IDs and account selection across the frontend.

1 commit · 188 files

The backend now enforces organization-scoped authorization consistently across domain actions, including DNS updates and domain export requests. It adds holder consent and identity-link validation, scope-filtered audit history, and approval-time binding of verified ownership data.

1 commit · 90 files

The project updates its runtime and dependency set, including Next.js and related libraries, and disables telemetry for production start scripts. It also adds an automated security scan workflow and adjusts MCP OAuth tests for correct server-side navigation behavior.

4 commits · 14 filesAdded +391Deleted −1,030

News agent delivery now avoids redundant X ingestion by following the configured schedule and adds tighter validation for research and analytics windows. Approved posts are held only when supported by explicit verbatim evidence pairs, with clearer failure reasons when research or checks fail.

1 commit · 11 filesAdded +367Deleted −38