Product Changelog

New Feature

• Run an agent on demand: Build an agent once, then run it interactively from a Polaris conversation whenever you need something done, instead of waiting for an event or a schedule. Type @ to pick the agent, or use the Run button in the Agent Builder.
• Runs as you: An on-demand agent acts as the person who runs it, using their connected accounts, their permissions, and the inputs they supply for that run, all recorded against their name.
• Per-run inputs: Define the inputs the agent asks for on every run, and choose which tool fields the runner fills in.
• Interactive: Watch each step stream live, answer the agent's questions mid-run, approve changes before they happen, and open a full execution log afterward.

To learn more, refer to the Run an Agent On Demand documentation.

New Feature

• Polaris Skills: Create reusable, named sets of instructions that Polaris applies whenever they are relevant, such as your naming conventions, your tone, or the checks that must always run. Keep a skill private, share it with specific people, or make it available to your whole organization.
• Two ways to create: Build a skill from the Skills page, or ask Polaris to draft one for you in a conversation.
• Deliberate or automatic: Run a skill on demand by typing / in the composer, or let Polaris apply it automatically when your request matches the skill's description.
• Sharing and governance: Control access with Restricted or organization-wide sharing, assign Can use or Can edit, and mark organization-wide skills as required so they reach everyone.

To learn more, refer to the Get Started with Polaris Skills documentation.

New Feature

• Build an agent from a Polaris conversation: Describe what you want in plain language, and Polaris interviews you for the few things it needs, selects matching tools, writes the instructions, and creates a draft agent. Review and publish it from the Agent Builder — the fastest way from an idea to a working agent.

To learn more, refer to the Build an Agent in Polaris documentation.

New Feature

• Polaris Full Screen Mode: Open Polaris in a full-screen workspace — your chats, Skills, and Agents in one place, with more room to work.

This release brings more control over asset metadata, a way to clear out duplicate images, AI metadata written for you at upload, and a listing that remembers how you left it.

• Select type user-defined field: A Select field captures metadata from a list you define instead of free text, so values stay consistent and filter together. Show the choices as a dropdown, radio buttons, or checkboxes. Read more in Configure Select Fields.
• Duplicate Management for Images: Identical images are found in the background and grouped into sets. Open Duplicate Management from the asset listing to review the copies and keep the one you want. Nothing is flagged during upload. Read more in Manage Duplicate Images.
• AI metadata at upload: Configure a space to write AI tags and alt text into assets as they arrive. Anything the uploader supplies is never overwritten. Read more in Auto-Populate AI Metadata at Upload.
• AI suggested metadata in the API: Read the tags, alt text, objects, colors, and safety flags Contentstack generates for an image, and enrich content from your own systems. Read more in the Assets API documentation.
• Persistent user preferences: The asset listing remembers how you left it. Your page size, view mode, column layout, sort order, and filters are saved to your account rather than to the browser, so they follow you to any browser or device you sign in from.

We are looking forward to seeing what you build with these.

Note: Assets is a paid add-on and it may not be enabled for your organization. Duplicate Management is being rolled out gradually and may not be available to you yet. To purchase Assets or to ask about availability, contact our support team.

Contentstack is no longer just for the world’s best enterprises. Starting today, anyone can now create a Contentstack account and start using our leading headless CMS and other powerful apps to build their next great project. Simply visit the Sign Up link to get your organization set up, create your first stack, and run your subscription yourself from the Subscription page in your account settings, click your avatar, then Manage Subscription. This new offering is a perfect fit for smaller teams and individual builders who want to get started and grow at their own pace. What will you build with Contentstack?

Key Features:

• Create your account and start building: Sign up, verify your email address with a verification code, set your password, and create your first stack. Compare subscription options on our pricing page before you choose.
• Pick the subscription that fits: Start on Free and move up to Build or Growth as your needs change, with a clear view of what you are charged today before you confirm.
• Extend with add-ons: Add capacity to your subscription and adjust quantities as your usage grows.
• Track usage against your limits: The Usage Overview shows where your organization stands on every resource your subscription includes, so you know before you run out rather than after.
• Turn on extra usage: Keep working past an allowance instead of hitting a hard stop, with the option to cap how far it goes.
• Keep your billing details current: Update your card, billing address, and invoices in the billing portal whenever you need to.
• Change course at any time: Cancel your subscription from the Subscription page on your own terms.

To see the limits that apply to your organization, check your own Subscription page, it is always current for you.

Check our documentation for full details, or contact support for more information.

You can now run a bulk action on every entry in the entries list instead of only the entries on the current page. If a search returns 200 entries, you no longer select 100, run the action, and repeat it for the rest, select all records and run it once on all 200.

From the entry list page, you can:

• Select beyond the current page: Select the checkbox in the table header, then click Select all {number} entries in the banner that appears.
• Run the bulk actions you already use: Publish, unpublish, delete, and add to release all work exactly as before, only the size of the selection changes.
• Filter first, or do not: Apply search, filters, or a view to narrow what you select, or select everything the list shows.
• Review before you act: Click Show Selected to see every entry in the selection, and Clear selection to start over.

Behavioral change: The limit query parameter now accepts positive integers only.

Passing limit=0, an empty value, or a non-numeric value now returns the default number of records for that resource instead of all matching records. Previously, limit=0 returned every matching record, which could cause requests on large stacks to time out.

If your integration relies on limit=0 to fetch a full result set, update it to paginate:

• Set an explicit limit and page through the results using the skip parameter.
• Pass include_count=true to retrieve the total number of records available, and continue paginating until you have retrieved that total.

By default, the response limit is 100 records per request. To retrieve the complete result set, specify an appropriate limit and use the skip parameter to retrieve each subsequent batch. For more information, refer to the Limit section.

We have introduced support for multiple identity providers in a Contentstack organization. An organization can hold up to five SAML 2.0 identity providers, each set up as a separate connection with its own certificate, session policy, role mapping, and login URL.

This suits organizations whose people do not all live in the same identity system: a parent company and an acquired subsidiary running separate tenants, employees and contractors on different providers, or a staged migration between providers where both need to stay live during the cutover.

Key Features:

• Up to five connections per organization: Each connection is configured, enabled, and disabled independently from the Single Sign-On page in Administration.
• A login URL for every connection: Each connection carries its own SSO ID, which forms its login URL. Share each connection's login URL with the users who should sign in through it.
• A primary connection: Any enabled connection can be marked as primary. The primary connection supplies the SSO login link that goes out in organization invitation emails.
• Per-connection session and role settings: Session time-out and IdP role mapping apply per connection, so the roles a user receives come from the connection they signed in through.
• Metadata import: An identity provider's sign-in URL and signing certificate can be imported from a metadata URL or an XML file, rather than entered by hand.
• Service provider metadata export: Contentstack's SP metadata for any connection downloads as an XML file, ready to pass to an identity provider team.
• Organization-wide strict SSO: Strict SSO applies across the organization. If any active connection has it enabled, no user can sign in without SSO.

Existing SSO setups are unaffected. A connection that is already in place keeps its SSO ID, login URL, and Assertion Consumer Service URL, so SAML applications configured against it continue to work without changes.

Note: Support for multiple identity providers might not be available by default. To enable it for your organization, contact our support team.

Check our Configure Multiple Identity Providers documentation for full details, or contact support for more information.

Embedded entries and assets in the JSON Rich Text Editor (RTE) now resolve to their current metadata, so your rendered content no longer serves values captured when the entry was last saved. You can also retrieve nested embedded items in a single request.

With this release, you can:

• Get current metadata automatically: Utils SDKs resolve asset and entry data from _embedded_items, so current metadata appears without stale or unpublished values leaking through, including for link-type references.
• Resolve nested items in one call: pass include_embedded_items[]=RECURSIVE to resolve entry-to-entry-to-asset chains without a separate request per level.
• Control the depth: use embedded_items_depth to limit how many levels are resolved, up to a maximum of 5.

Security Fix:

• Updated the jsoup dependency to version 1.23.2 to resolve a resource allocation vulnerability (CVE-2026-75140).

Security Fix:

• Updated the jsoup dependency to version 1.23.2 to resolve a resource allocation vulnerability (CVE-2026-75140).

Security Fix:

• Updated the jsoup dependency to version 1.23.2 to resolve a resource allocation vulnerability (CVE-2026-75140).

Bug Fixes:

• @contentstack/cli-utilities (2.0.1):
  • Fixed an intermittent Maximum call stack size exceeded crash during OAuth token refresh caused by synchronous re-entrant recursion.
  • Fixed unbounded retries on persistent 401 errors. It now retries once and exits with a clear error message.
• @contentstack/cli-cm-import (2.0.1):
  • Clarified the global field reference-field warning message shown when importing a global field.
• @contentstack/cli-bulk-operations (2.1.0):
  • Fixed an issue where assets were being published to all environments instead of only the scoped targets.
• @contentstack/cli-cm-bootstrap (2.0.1):
  • Fixed bootstrap failing silently on non-2xx tarball responses, hidden spinner errors, and imprecise branch-not-found error messages.
• contentstack-cli-tsgen (5.1.0):
  • Fixed the tsgen GraphQL type-generation command failing with a generic schema error instead of surfacing the real underlying error.
• types-generator (3.10.5):
  • Fixed the tsgen GraphQL type-generation command failing with a generic schema error instead of surfacing the real underlying error.

We’ve made the Assets API available, so everything you set up in Contentstack Assets can now be driven from your own systems. The interface is built for a person working on one asset at a time; the API applies the same structures across thousands of them, on a schedule, or as a step in your release pipeline.

What you can do with it

• Provision spaces and workspaces: Create spaces per brand, team, or region, add isolated workspaces for a campaign, and set their locales, so onboarding a new team becomes a script rather than a checklist.
• Model assets around your business: Define asset types and custom metadata fields through the API, then write values onto assets in bulk. A taxonomy change can be rolled out across an entire library in one pass.
• Localize without re-uploading: Manage localized asset variants and workspace locales programmatically, so opening a new market does not mean rebuilding a library by hand.
• Automate governance: Read an asset’s full version history, fetch any specific version, and name versions so your audit trail matches your release process.
• Enrich assets for interactive experiences: Set visual markups such as hotspots and bounding boxes, and apply image transformations, as part of your publishing pipeline instead of asset by asset.
• Search the way the interface does: Combine nested filters, restrict a query to specific fields or spaces, page through results, work locale-aware with fallbacks, and use AI-assisted suggestions.
• Work in bulk, and recover: Move, delete, and trash assets in bulk, and restore the ones you need from the bin.
• Manage access: Invite users to a space and set their roles, so permissions stay in step with your own systems of record.

The API is available in AWS North America, AWS Europe, Azure North America, and GCP North America. Authenticate with a Contentstack authtoken or an OAuth access token, and pass the x-cs-api-version header with a value of 4 on every request.

We’re looking forward to seeing what you build on it. Check our Assets API documentation for full details.

Note: Assets is a paid add-on and it may not be enabled for your organization. To purchase it, contact our support team.

You can now update the same text across multiple entries in a single operation from the entry list page with Find and Replace. Filter and select your entries, search for a text string, then replace it or remove it in a single operation.

From the entry list page, you can:

• Match text your way: choose Contains, Matches, or Exactly matches to control how strictly your search term is compared.
• Replace or remove: enter replacement text, or leave it empty to delete every matching occurrence.
• Preview before you apply: review each match by entry, locale, and field, including the number of matches and the resulting state, then fine-tune your selection.
• Add to a release: bundle the successfully replaced entries into a new or existing release for publishing.

New Features:

• Added publish and unpublish support for taxonomies and terms.
• Added localize and unlocalize support for taxonomies and terms.

You can now assign names to specific versions of your variant entries, the same way you name versions of a base entry. Naming a version makes it easy to identify a key change or milestone at a glance, instead of scanning version numbers.

With this update, you can:

• Name an entry variant version directly from the version history; the name is saved without a separate save step.
• Find the right version faster using meaningful names instead of version numbers.
• Manage names through the API with new Content Management API calls to set, list, and remove variant version names.

Check our documentation for more details, or contact support for more information.

We've added environment selection to branch creation, so you can choose which environments to clone into a new branch instead of cloning all environments by default.

In the Create New Branch modal, or through the Content Management API, you decide whether a branch clones all environments, none, or a specific set.

When you create a branch, you can:

• Clone specific environments: Select only the environments the branch needs under Specific, then Select Environments.
• Clone all or none: Choose All to clone every environment from the source branch, or None to clone no environments.
• Control cloning from the API: Pass publish_config.environments on the create-branch endpoint to set the same behavior programmatically.

Note
Specific is the default in the UI, so the Create button stays disabled until you select at least one environment. The API differs: it clones all environments by default when publish_config is omitted, so existing integrations are unchanged.

Check our documentation for full details, or contact support for more information.

Bug Fix:

• Upgraded the json gem to 2.21.2 to fix a high-severity use-after-free vulnerability (CVE-2026-71847) in the streaming JSON parser.

Bug Fix:

• Resolved a security vulnerability.

Bug Fix:

• Upgraded org.jsoup:jsoup to 1.23.1 to fix a Cross-site Scripting (XSS) vulnerability.

Bug Fixes:

• Fixed a crash when logging requests whose URLs contain percent-encoded characters, caused by the log message being passed to the system logger as a format string.
• Fixed a runtime error when using the cacheThenNetwork cache policy with async/await. The policy now resolves once and falls back to cache-else-network.
• Fixed a per-request networkElseCache cache policy being ignored in favor of the stack-level policy when a transport error occurred.
• Fixed requests hanging indefinitely when a response returned neither data nor an error. These now fail with an invalid HTTP response error.

Bug Fix:

• Resolved a security vulnerability.

New Features:

• @contentstack/cli (2.0.0):
  • Introduced a guided warning and install prompt for missing launch:* or cm:entries:migrate-html-rte commands instead of a bare "command not found" error, making plugins opt-in at GA.
• @contentstack/cli-asset-management (1.0.0):
  • Introduced a new plugin providing full Asset Management 2.0 (AM 2.0) API support, covering spaces, workspaces, fields, and asset types.
  • Added OAuth authentication support for AM.
• @contentstack/cli-bulk-operations (2.0.0):
  • Consolidated 14 separate bulk-publish commands into unified commands (cm:stacks:bulk-entries, cm:stacks:bulk-assets).
  • Added the new cm:stacks:bulk-taxonomies command for bulk publishing or unpublishing taxonomy terms.
  • Integrated the Progress Manager UI.
  • Configured NRP to be used on all publish/unpublish calls.
• @contentstack/cli-cm-migrate-rte (2.0.0):
  • Made the RTE migration available as a separate, opt-in plugin via csdx plugins:install @contentstack/cli-cm-migrate-rte.
• @contentstack/cli-auth (2.0.0):
  • Added the new auth:tokens:list command for listing saved tokens and introduced the auth:tokens namespace as help.
• @contentstack/cli-utilities (2.0.0):
  • Added a new readGlobalFieldSchemas utility function.
  • Introduced a retry mechanism with backoff for transient HTTP errors (network / 429 / 5xx), used across AM and export/import.
  • Implemented a strategy pattern to support Progress Manager and Summary Manager rollout across plugins.
  • Registered bulk-operations as a Progress-Manager-supported module.

Enhancements:

• @contentstack/cli-cm-export (2.0.0):
  • Added AM 2.0 export support, featuring automatic detection via org plan-check to correctly skip assets under a management token.
  • Updated the export structure to output a flat directory.
  • Main branch is exported by default for branch enabled stack when branch flag is empty.
  • Changed Global fields to export as individual files (one per item) rather than a single combined file.
  • Removed the content type schema JSON export.
  • Added concurrency support for faster AM exports and improved logging.
  • Integrated Progress Manager and Summary Manager across stack, assets, environments, taxonomies, extensions, global fields, locales, personalize, and variant-entries export.
  • Removed deprecated short flags in favor of long-form only (--data-dir, --stack-api-key, --alias, etc.).
• @contentstack/cli-cm-import (2.0.0):
  • Added AM 2.0 import and asset publishing support.
  • Updated taxonomies to re-publish automatically after import, and added a new --skip-taxonomy-publish flag to allow review before publishing.
  • Integrated Progress Manager and Summary Manager.
  • Removed deprecated short flags in favor of long-form only.
• @contentstack/cli-cm-import-setup (2.0.0):
  • Added AM 2.0 support and Progress Manager integration.
• @contentstack/cli-cm-seed (2.0.0):
  • Refined the interactive stack-seed picker to display a curated list of 3 official starter repos instead of searching GitHub live for a faster, more reliable experience.
• @contentstack/cli-bulk-operations (2.0.0):
  • Removed the --api-version flag from cm:stacks:bulk-entries and cm:stacks:bulk-taxonomies as the NRP header is used by default.
• @contentstack/cli-config (2.0.0):
  • Added optional --cs-assets and --auth-api flags to config:set:region for setting custom Contentstack Assets and Auth API endpoints.
  • Improved error handling for rate-limit failures and region-endpoint resolution.
• @contentstack/cli-utilities (2.0.0):
  • Consolidated session log files into a single folder per run.
  • Renamed the console log config key internally to showConsoleLogs.
• @contentstack/cli-variants (2.0.0):
  • Ensured entry variants and variant groups correctly respect the --branch flag on export and import instead of always operating against the default branch.
• @contentstack/apps-cli (2.0.0):
  • Added support for -k as a shorthand for --stack-api-key.
• contentstack-cli-tsgen (5.0.0):
  • Renamed the --token-alias flag to --alias.
  • Removed -o, -p, and -d short flags in favor of long-form flags only.
• contentstack-cli-content-type (2.0.0):
  • Removed deprecated flags and ambiguous short characters across all six content-type:* commands (audit, compare, compare-remote, details, diagram, list).
• @contentstack/cli-migration (2.0.0) & @contentstack/cli-external-migrate (2.0.0):
  • Cleaned up additional short-character flags in favor of long-form flags only.
• @contentstack/cli-cm-regex-validate (2.0.0):
  • Added new message strings to the plugin.
• Cross-Plugin Enhancements:
  • Implemented clearer, conditional success/warning messaging and consolidated end-of-run failure summaries across export, import, and bulk operations.

Bug & Security Fixes:

• @contentstack/cli (2.0.0):
  • Fixed an issue where Ctrl+C during an interactive prompt threw an uncaught error; it now exits cleanly with code 130.
  • Resolved several false-positive hardcoded-secret findings flagged during a routine security scan.
• @contentstack/cli-auth (2.0.0):
  • Fixed a two-factor authentication (2FA) login issue.
• @contentstack/cli-utilities (2.0.0):
  • Fixed a race condition in OAuth token refresh where concurrent refresh attempts could conflict.
• @contentstack/cli-cm-clone (2.0.0):
  • Resolved a prompt issue during clone import.
  • Fixed a path-resolution bug that broke when the CLI was installed under a directory path containing /lib/ (e.g., Node's own lib/node_modules).
• @contentstack/cli-cm-export (2.0.0) & @contentstack/cli-cm-import (2.0.0):
  • Fixed a null/undefined crash affecting multiple export/import modules (custom roles, environments, extensions, labels, locales, taxonomies, webhooks, and workflows).
  • Resolved a duplicate prompt for the marketplace app encryption key during export.
  • Corrected the taxonomy and extension mapper path used during import, which previously pointed to the wrong directory.
• @contentstack/cli-cm-export-to-csv (2.0.0):
  • Fixed pagination when fetching organization users for organization owners with a large number of users.
• @contentstack/cli-cm-seed (2.0.0):
  • Fixed a directory-change ordering bug that could break a nested import during seeding.
• @contentstack/cli-audit (2.0.0):
  • Fixed the prompt shown when auditing a stack with no global fields.
• @contentstack/cli-variants (2.0.0):
  • Fixed a content-type linking failure when a personalize experience had no audience.
• @contentstack/cli-external-migrate (2.0.0):
  • Resolved a path traversal vulnerability in the Contentful migration adapter.

Bug Fixes:

• Upgraded @contentstack/core to ^1.5.1 to resolve MODULE_NOT_FOUND and "expression is too dynamic" build failures in bundlers such as Next.js Turbopack and webpack, caused by a dynamic require() call in the keep-alive agent setup. It now uses statically analyzable require('http') and require('https') calls.
• Guarded keep-alive agent creation for native ESM environments, where require is undefined, to prevent a ReferenceError.
• Added a package browser field mapping http/https so browser bundles resolve them cleanly.

Bug & Security Fixes:

• Updated dependencies in datasync-asset-store-filesystem (v2.3.8) to resolve security vulnerabilities.

New Feature:

• Added Taxonomy CDA support: fetch and query published taxonomies and terms with hierarchy traversal, localization, and fallback.

Bug & Security Fixes:

• Updated vulnerable dependency versions in contentstack.management.aspnetcore.csproj and test project files to resolve security vulnerabilities.
• Suppressed static-analysis alerts for hardcoded credentials in test files (integration and unit tests).

Bug & Security Fix:

• Resolved security vulnerabilities.

Bug & Security Fixes:

• Bumped @contentstack/core dependency to ^1.5.0, introducing the following stability improvements:
  • Added default retry logic for transient network-level errors (e.g., ECONNABORTED, ETIMEDOUT, ECONNRESET, EPIPE, EAI_AGAIN) when there is no HTTP response.
  • Implemented distinct classification for request timeouts instead of returning a generic UNKNOWN_ERROR.
  • Added default httpAgent and httpsAgent keep-alive connection agents in Node environments.

Contentstack now supports organization-specific HMAC signing for webhook payloads, giving you greater security and control over webhook verification. Contentstack's earlier webhook signing model used one platform-wide certificate across all payloads. HMAC signing gives each organization its own secret key, generated and managed from Administration > Security Configuration > HMAC Signing, and used to sign webhook payloads with the HMAC-SHA256 algorithm.

What's New:

• Isolated signing keys: Each organization signs with its own secret key, ensuring that keys are isolated across organizations. Consumers can verify that requests originating from Contentstack were not altered in transit.
• Self-service key rotation: Regenerate your secret key any time directly from the UI, without contacting support.
• Zero-downtime rotation: When you regenerate a key, choose how long the previous key stays valid, from immediately up to 24 hours. During the grace period, Contentstack signs payloads with both keys so existing consumers continue working while you update them.
• Per-webhook control: Choose the signing method for each webhook using the new Request Signing Method field to gradually adopt HMAC while other webhooks continue using the default Contentstack certificate.
• Secure by design: Secret keys are encrypted and are never exposed after creation. Each payload includes a timestamp to help your application detect and reject replayed requests.

Users with the Owner, Admin, Security Manager, or a custom role with the required permissions can enable, regenerate, or disable HMAC signing.

Note: This feature might not be enabled by default. Contact our support team to get it enabled for your organization.

Learn how to enable HMAC signing, rotate your key, and verify signatures in our HMAC Signing documentation.

We’ve added new enhancements to the Brightcove app:

• Multiple Configuration Support: You can now set up multiple Brightcove accounts using separate configurations.
• Video ID Search: Search and add videos by video ID directly from the selector page.
• Advanced Configuration Object: Added support for configuration objects in advanced settings for the Custom and JSON RTE fields for easier setup and management.
  
  • Branch-specific Configuration Support: Set branch-specific parameters to scope asset selection to a particular branch.
  • Locale-specific Configuration Support: Set locale-specific parameters to tailor asset selection by language.

To learn more, refer to the Brightcove App Installation Guide.

Bug & Security Fixes:

• Bumped setuptools minimum version from >=80.0.0,<82.0.0 to >=83.0.0 to resolve a security vulnerability finding.

Bug & Security Fixes:

• Upgraded setuptools to 83.0.0 to resolve an Improper Unicode Encoding Handling security issue.

Bug & Security Fixes:

• Upgraded black to >=26.3.1 to fix a high-severity Directory Traversal vulnerability.
• Upgraded pip to >=26.1.2 to fix Arbitrary File Upload, Arbitrary Code Injection, and Directory Traversal vulnerabilities.
• Upgraded setuptools to >=83.0.0 to fix an Improper Unicode Encoding Handling vulnerability.

Enhancements:

• Updated the contentType(...).entry(...).variants(variantUidOrUids, branchName?) method to accept an optional branchName as the second argument, which sets the CMA branch header for that variant's scope. The first argument supports either a single variant UID string or an array of UIDs (comma-separated in the request path). Omitting the branch name preserves the previous behavior.
• Added variants(uid).publish() and variants(uid).unpublish() methods, which call the entry publish and unpublish endpoints by nesting the variant payload under the entry.
• Updated the publish() and unpublish() methods (for both entries and entry variants) to accept optional headers and parameters, which are merged directly into the underlying HTTP request.
• Added unit and sanity API tests for entry variants with an explicit branch, as well as for the new variant publish and unpublish methods.

New Features:

• Added branch support in entry variants.

Enhancements:

• Updated the variants() method on Entry and Entries to support an optional branch name as the second argument. When provided, the branch is sent as a request header alongside x-cs-variant-uid.
• Ensured existing variants(uid) and variants(uids) calls remain backward compatible.
• Added unit and API tests to support variant and branch requests.

New Features:

• Added branch support in entry variants.

Enhancements:

• Added support for entry variants with optional branch scoping via the variants() method on Entry and Query.

Enhancements:

• Added the variants(variant_uids, branch_name) method to Contentstack::Entry and Contentstack::Query to fetch entry variants with optional per-request branch scoping. These requests now send the x-cs-variant-uid header (as comma-separated UIDs) and respect either the stack-level or per-call branch settings.

Enhancements:

• Made the launch plugin opt-in, it is no longer bundled with the CLI. Install it using csdx plugins:install @contentstack/cli-launch to use launch commands in CLI (v2.0.0-beta.30).
• Added the readGlobalFieldSchemas utility function in cli-utilities (v2.0.0-beta.12).
• Automatically refreshed region configurations set on older CLI versions with the latest endpoint data upon the next command execution, eliminating the need to manually run config:set:region in cli-utilities (v2.0.0-beta.12).
• Updated the config:set:region command to store the full endpoint set for the selected region, making any future Contentstack endpoints available without requiring a command re-run in cli-config (v2.0.0-beta.15).
• Switched global fields to a per-file export/import format in the following CLI plugins:
  • cli-export (v2.0.0-beta.25)
  • cli-import (v2.0.0-beta.25)
  • cli-audit (v2.0.0-beta.16)
  • cli-import-setup (v2.0.0-beta.19)
• Added global field rule handling to content types in cli-import (v2.0.0-beta.25).
• Added addHeader to the publish-entries chain in cli-import (v2.0.0-beta.25).
• Enhanced the field rules audit to include global fields in cli-audit (v2.0.0-beta.16).
• Enhanced Asset Management (AM) with asset publishing support in cli-asset-management (v1.0.0-beta.8).
• Forced api_version=3.2 on all entry and asset publish/unpublish call sites in the following CLI plugins:
  • contentstack-external-migrate (v2.0.0-beta.4)
  • cli-bulk-operations (v2.0.0-beta.5)
• Hardened the NRP header on taxonomy publish and fixed include-variants validation in cli-bulk-operations (v2.0.0-beta.5).

Bug & Security Fixes:

• Reconciled completed progress bar counts with processed items in cli-utilities (v2.0.0-beta.12).
• Resolved an issue where resolving the org plan or auth host would fail hard when not explicitly configured, and successfully derived it from the existing region host as a fallback in cli-utilities (v2.0.0-beta.12).
• Fixed context setting issues in cli-command (v2.0.0-beta.11).
• Fixed an issue to add global fields Field Validation Rules (FVRs) in exports in cli-export (v2.0.0-beta.25).
• Fixed an issue by collecting all locales and deduplicating asset UIDs in the bulk payload in cli-bulk-operations (v2.0.0-beta.5).
• Updated dependency versions in the following CLI plugins:
  • cli-auth (v2.0.0-beta.17)
  • cli-command (v2.0.0-beta.11)
  • cli-config (v2.0.0-beta.15)
  • apps-cli (v2.0.0-beta.5)
  • cli-cm-bootstrap (v2.0.0-beta.25)
  • cli-cm-branches (v2.0.0-beta.11)
  • cli-cm-regex-validate (v2.0.0-beta.4)
  • contentstack-cli-tsgen (v5.0.0-beta.4)
  • cli-cm-clone (v2.0.0-beta.26)
  • contentstack-cli-content-type (v2.0.0-beta.3)
  • cli-cm-export-to-csv (v2.0.0-beta.12)
  • cli-cm-migrate-rte (v2.0.0-beta.10)
  • cli-migration (v2.0.0-beta.17)
  • cli-cm-export-query (v2.0.0-beta.9)
  • cli-cm-seed (v2.0.0-beta.25)
  • cli-variants (v2.0.0-beta.20)

Enhancements:

• CLI (v1.66.0):
  • Added groundwork for organization plan-based feature gating for CLI commands (currently inactive until individual commands opt in).
  • Introduced an auto-refresh for endpoints on command pre-run.
• Added support for global field validation rule for the following plugins:
  • cli-audit (v1.20.0)
  • cli-export (v1.26.0)
  • cli-import (v1.34.0)

Updated the config:set:region command to store the full endpoint set for the selected region. Any future Contentstack endpoint added will now be available without requiring a command re-run.

Region configurations set on older CLI versions are now automatically refreshed with the latest endpoint data upon the next command execution, eliminating the need to manually run config:set:region.

@contentstack/cli-cm-clone (v1.21.10)@contentstack/cli-cm-export-query (v1.0.6)@contentstack/cli-cm-seed (v1.15.9)@contentstack/cli-auth (v1.8.5)@contentstack/apps-cli (v1.7.3)@contentstack/cli-cm-bootstrap (v1.19.8)@contentstack/cli-cm-branches (v1.8.4)@contentstack/cli-cm-bulk-publish (v1.12.2)contentstack-cli-tsgen (v4.10.2)@contentstack/cli-cm-export-to-csv (v1.12.7)@contentstack/cli-external-migrate (v1.0.0)@contentstack/cli-cm-import-setup (v1.8.7)@contentstack/cli-migration (v1.12.5)@contentstack/cli-variants (v1.6.1)

Bug & Security Fixes:

• cli-utilities (v1.20.0):
  • Resolved an issue where resolving the org plan or auth host would fail hard when not explicitly configured. It now successfully falls back to deriving from the existing region host.
• cli-command (v1.8.6):
  • Fixed issues related to context setting.
• Security Fixes: Resolved security vulnerabilities in the following plugins:
  • cli-bulk-operations (v1.2.3)
  • cli-cm-regex-validate (v1.0.2)
  • contentstack-cli-content-type (v1.5.3)
  • cli-cm-migrate-rte (v1.7.3)

We’ve enhanced Live Preview so that teams editing the same entry now share a collaborative draft preview. 

Previously, each user had a separate preview and saw only their own unsaved changes. Now, when multiple users edit the same entry, everyone sees the same content as it changes.

With Collaborative Live Preview, you can:

• See collaborators’ edits in your preview as they work on the same entry.
• Stay in sync automatically, your own changes appear instantly, and others’ changes appear after a brief delay, so the preview does not reload on every edit.
• Preview every field, since collaborative preview applies to all fields in the entry.

Note: Collaborative drafts are not yet available in Visual Editor.

Refer to the Collaborative Drafts in Live Preview documentation for full details, or contact support for more information.

Bug Fix:

• Fixed an issue where the previously published 2.0.0 package did not include the EmbeddedObjectConverter. Bumped the package version to 2.0.1 and republished to ensure the published package matches the source code unambiguously.

Bug & Security Fix:

• Bumped the @contentstack/utils dependency from ^1.4.1 to ^1.9.1.

New Features:

• Added support for passing an optional branch parameter to the .Variant() method in both the Entry and Query classes to support Entry Variants branching.
• Implemented an automatic fallback to the Stack's configured branch or "main" if the branch parameter is null or empty.
• Added comprehensive unit and integration tests for the Entry and Query variant branch logic.

New Features:

• Added optional branch support for entry variants in the Entry.variants() and ContentType.variants() methods.

New Features:

• Added branch override support for Entry Variants. The Entry.Variant(uid?, branchUid?) method now accepts an optional branchUid to target a specific branch for variant calls (Find, Create, Update, Fetch, Delete) by overriding the branch request header. If null or empty, it falls back to the Stack's configured branch.
• Added publish and unpublish support for Entry Variants via EntryVariant.Publish/PublishAsync and EntryVariant.Unpublish/UnpublishAsync. These methods match the PublishUnpublishDetails contract used by Entry.Publish/Unpublish and respect the branch override.

New Features:

• Added publish and unpublish support for Entry Variants via the entry publish/unpublish endpoints, including documented payloads for entry.variants and optional entry.variant_rules on publish.
• Added optional stack branch support via the branch request header for Entry Variants. Entry.variants() now accepts no arguments, a branch UID only, or (variant_uid, branch) (use variants(variant_uid, None) when targeting a variant without a branch).

New Feature:

• Added branch support in entry variants.

New Feature:

• Added branch support in entry variants.

Enhancements:

• cli-bulk-operations:
  • Merged the cm:stacks:bulk-am-assets command into cm:stacks:bulk-assets. All asset functionality is now supported directly by cm:stacks:bulk-assets, removing the need for a specific assets command.
  • Integrated the Progress Manager UI into all cm:stacks:bulk-* commands, ensuring bulk operations now display a clean header, a live summary, and per-command Module Details.
• Upgraded dependencies for the following plugins:
  • cli-utilities (v2.0.0-beta.11)
  • cli-bulk-operations (v2.0.0-beta.4)
  • cli-auth (v2.0.0-beta.16)
  • cli-command (v2.0.0-beta.10)
  • cli-config (v2.0.0-beta.14)
  • apps-cli (v2.0.0-beta.4)
  • cli-asset-management (v1.0.0-beta.7)
  • cli-audit (v2.0.0-beta.15)
  • cli-cm-bootstrap (v2.0.0-beta.24)
  • cli-cm-branches (v2.0.0-beta.10)
  • cli-cm-regex-validate (v2.0.0-beta.3)
  • contentstack-cli-tsgen (v5.0.0-beta.3)
  • cli-cm-clone (v2.0.0-beta.25)
  • contentstack-cli-content-type (v2.0.0-beta.2)
  • cli-cm-export (v2.0.0-beta.24)
  • cli-cm-export-to-csv (v2.0.0-beta.11)
  • cli-external-migrate (v2.0.0-beta.3)
  • cli-cm-import (v2.0.0-beta.24)
  • cli-cm-import-setup (v2.0.0-beta.18)
  • cli-cm-migrate-rte (v2.0.0-beta.9)
  • cli-migration (v2.0.0-beta.16)
  • cli-cm-export-query (v2.0.0-beta.8)
  • cli-cm-seed (v2.0.0-beta.24)
  • cli-variants (v2.0.0-beta.19)

New Feature:

• Added support for Taxonomy Localization and Publishing.

Enhancement:

• Removed the locale parameter from Taxonomy.fetch() and Term.fetch().

New Feature:

• Added Taxonomy Publishing support to the Content Delivery SDK via stack.taxonomy().

Lytics CDP is now available from the App Switcher in your Contentstack organization. It's Contentstack's interface for managing behavioral data collection, audience building, and personalization activation across your digital properties, all connected directly to your CMS, Launch, and Personalize projects.

With Lytics CDP, you can:

• Collect behavioral data automatically: Install the JStag SDK on your front end, or enable it with a single toggle if your site runs on Contentstack Launch, to start capturing visitor events across web, mobile, and other digital touchpoints.
• Connect your Contentstack products in one place: Link your CMS stacks, Launch projects, and Personalize projects to a Lytics project so audience data flows automatically, with no custom middleware or data pipelines required.
• Activate real-time personalization: Use live visitor behavior and audience membership, authored on the Lytics platform, to determine which content variant a visitor receives in Personalize.
• Manage everything at the project level: Create and manage Lytics projects, collaborators, and connections directly in Contentstack, with access controlled through your existing organization roles.

Lytics CDP works alongside the Lytics platform (app.lytics.com), where audience authoring, profile inspection, and data pipeline configuration still happen. You can jump straight into the Lytics platform from your project dashboard using the Manage Lytics button and single sign-on.

If you previously connected Lytics through the Data Activation Layer (DAL), your existing configuration, data, users, and settings have been automatically converted into a Lytics CDP project, and your Lytics audiences continue to work in Personalize without any changes.

What's New?

• Stack-Aware Tool Enrichment: Enabled by default, tool descriptions, enums, defaults, and argument guidance are now tailored to your connected Contentstack stack.
  • Enrichment support across CDA, CMA, CMA Extended, Launch, Brand Kit, Personalize, and Developer Hub tools
  • New x-enrich, x-enrich-group, and x-enrich-description metadata added to tool definitions, plus a machine-readable tools/enrich-profiles.json reference
  • CONTENTSTACK_MCP_ENRICH_TOOLS=false available to disable enrichment while still stripping raw enrichment metadata from exposed tool schemas
  • Optional stack context inputs: CONTENTSTACK_ENVIRONMENT, CONTENTSTACK_BRANCH, and CONTENTSTACK_LOCALE
• Taxonomy Term Reordering and Reparenting: A new move_a_term tool in the CMA group supports reordering and reparenting a taxonomy term via PUT /v3/taxonomies/{taxonomy_uid}/terms/{term_uid}/move. Previously, update_a_term only supported renaming — hierarchy changes were not possible through the server.
• Tool-Selection Evaluation Harness: A new OpenAI evaluation harness under eval/ measures MCP tool-selection behavior with enriched tool definitions.
• To learn more, refer to the Contentstack MCP Server documentation.

Enhancements:

• Upgraded dependencies for the following plugins:
  • cli-cm-export (v2.0.0-beta.23)
  • cli-cm-import (v2.0.0-beta.23)
  • cli-bulk-operations (v2.0.0-beta.3)
  • cli-cm-bootstrap (v2.0.0-beta.23)
  • cli-cm-clone (v2.0.0-beta.24)
  • cli-cm-export-to-csv (v2.0.0-beta.10)
  • cli-cm-import-setup (v2.0.0-beta.17)
  • cli-cm-seed (v2.0.0-beta.23)
  • cli-variants (v2.0.0-beta.18)
  • cli-asset-management (v1.0.0-beta.6)
  • cli-external-migrate (v2.0.0-beta.2)

Bug & Security Fixes:

• Fixed minor bugs in the cli-asset-management (v1.0.0-beta.6) plugin.

Documentation Updates:

• Updated the CLI v2 README with accurate installation instructions, command references, usage examples, and a clear summary of breaking changes from version 1.

Bug & Security Fixes:

• Fixed an issue where content types were not correctly linked during Personalize import when some experiences have no audiences in the following plugins:
  • cli-cm-import (v1.33.5)
  • cli-variants (v1.6.0)

Bug & Security Fix:

• Removed the unused keyring dependency to resolve a license-policy issue on the transitive secretstorage package.

Bug & Security Fix:

• Removed hardcoded password literals in unit tests to resolve CWE-798 (Use of Hardcoded Passwords) findings.

Managing multiple personalization experiences across campaigns, audiences, and content types can get complex fast. Experience Tags give you a flexible, searchable way to label, group, and filter your experiences in Personalize, so your team always knows what's running, why it exists, and where it belongs.

You can now create custom tags and attach them to any experience directly from the Personalize dashboard. Use tags to reflect your internal taxonomy, whether that's by region, campaign type, audience segment, lifecycle stage, or any other dimension that matters to your team.

With Experience Tags, you can:

• Stay organized at scale: Filter your experience list by one or more tags instantly. No more scrolling through dozens of experiences to find what's relevant.
• Apply multiple tags per experience: Support cross-functional views - for example, tag an experience as both "Holiday 2026" and "EMEA" simultaneously.
• Standardize naming across your team: Reuse tags created by any team member in the same project to keep your taxonomy consistent.
• Audit and govern with ease: Use tags as a lightweight governance layer when many experiences are running in parallel across teams or campaigns.

We have released Contentstack MCP Server v0.7.3, a fix for an issue affecting MCP clients that skip the tools/list step.

What's New?

• Added automations tool group with 9 tools: list_automation_projects, get_automation_project, create_automation_project, update_automation_project, delete_automation_project,list_automations, get_automation, trigger_automation, set_automation_active
• Added CONTENTSTACK_AUTOMATIONS_BASE_URL for dedicated automations infrastructure overrides

What's Fixed?

• Fixed server startup failure when invoked via npx @contentstack/mcp, caused by unresolved symlinks in entrypoint detection

To learn more, refer to the Contentstack MCP Server documentation.

We have released Contentstack MCP Server v0.7.3, a fix for an issue affecting MCP clients that skip the tools/list step.

What's New?

Lazy-loaded Tool Registry: The tool registry now lazy-loads on the first tools/call request when an MCP client invokes a tool without first calling tools/list.

• Some hosts (e.g., Cursor) cache tool schemas locally and may call a tool directly without listing available tools first.
• Previously, this triggered an error: Cannot read properties of undefined (reading).
• Tool calls now succeed regardless of whether tools/list was called beforehand.

To learn more, refer to the Contentstack MCP Server documentation.

We’ve added branch support for Variants, so you can build and test personalized content in an isolated branch before making it available on your main branch. Entry variants are no longer limited to the main branch.

You can now:

• Manage variant configuration on any branch: link content types and create entry variants in a development branch, isolated from main.
• Keep variant groups consistent: a variant group and its variants stay global across branches, while content type assignment and entry variants remain branch-specific.
• Test personalization in isolation: onboard to Personalize or run an A/B test without affecting production content on main.
• Merge variant groups across branches: merge variant group configuration from a compare branch into a base branch using the Content Management API, with five merge strategies.

Note: Merging variant groups is available through the API. Workflow status is not included in the merge.

Check our documentation for full details, or contact support for more information.

We are excited to announce the MCP Client tool in Agent OS Agents, enabling you to connect your agents to any remote Model Context Protocol (MCP) server and use its tools automatically, no custom integration code required.

MCP Client Tool: A new tool available in the Agent OS agent builder that connects your agent to any remote MCP server over HTTPS.

• Connect once and your agent gains access to all tools the server exposes, or the specific subset you approve.
• Support for two authentication methods: Header-based authentication (API key or token) and OAuth 2.1 (sign-in via provider consent screen).
• OAuth supports both automatic setup via Dynamic client registration and manual setup for providers that require a registered OAuth app.
• Allowed Tools selector lets you either allow the AI to choose freely from all available tools, or lock it down to only the specific tools you pick.
• Connections are reusable, a saved MCP Client connection can be attached to multiple agents from the Connected Apps page.

To learn more, refer to the MCP Client: Connect Remote Tools documentation.

Entries that use custom URL patterns now resolve in Visual Editor the same way they do in Live Preview, using the same configuration with no additional setup.

With this update, you can:

• Preview custom URL entries directly on the Visual Editor canvas.
• Browse pages from custom URL content types in the URL bar, with each entry’s resolved URL shown.
• Create pages for custom URL content types without a URL field.
• Resolve missing values from a screen that lists the fields a pattern needs, with a link to each field in the form.

For sites that use custom URLs, developers set the page context (Live Preview Utils SDK 4.4.4 or later) so the Start editing button opens the correct entry. See the SDK config reference for implementation details.

Note: Custom Preview URLs is available on select plans.

Learn more in the documentation, or contact support for more information.

Here’s an overview of the latest updates in Contentstack Launch:

• Response Mode Selection for Environments:

  You can now choose how your environment delivers responses with the new Response Mode setting — Streaming, which delivers response chunks in real time as they are generated, or Buffered, which displays output only after the entire response has been generated. This replaces the earlier Enable Streaming Responses toggle, giving you explicit control over response delivery and making the default behavior (Buffered) visible at a glance. A Response Mode column is now available in the Environments list table, letting you view each environment's response mode at a glance.

  Learn more in Environments.

• Project and Environment Selection in Log Targets:

  You can now select specific project and environment combinations when setting up a Log Target, and choose which type of logs — Server Logs, Traffic Logs, or both — to forward to each destination. Alternatively, you can choose to export logs for all current and future projects and environments within your organization. This gives you precise control over what data reaches each monitoring destination, reducing noise and keeping observability costs in check.

  Learn more in Log Targets.

What's New?

We are excited to announce Contentstack MCP Server v0.7.2, introducing OAuth scope declarations across all 196 tools in the nine /<group>/tools endpoints, giving the served JSON a single source of truth for tool-to-scope mappings used by the OAuth consent flow and runtime authorization.

Each tool entry now includes a scopes array alongside group, making the served JSON the authoritative source for the tool-to-scope mapping consumed by the OAuth consent flow and runtime authorization checks. Scopes are derived deterministically from (group, mapper.method, mapper.apiUrl, name) via src/utils/scopes.ts; run npm run generate:scopes to regenerate them as needed. Note that cda and lytics resolve to an empty array ([]), since these use delivery-token and Lytics-token authentication rather than OAuth.

This release includes two improvements to the entry editing experience.

Regex Validation for URL fields

You can now apply the Validation (Regex) and Validation Error Message properties to URL fields to control the URL format that content managers can enter. Previously, these options were available only for Single Line Textbox and Multi Line Textbox fields. With this change, you can:

• Define a custom regex pattern for the accepted URL format.
• Show a custom validation message when an entered value doesn't match the pattern.

Refer to the URL field documentation for details.

Live Preview Panel Now Stays Open in the Entry Editor

The Live Preview panel now follows the standard entry editor panel behavior, making navigation more predictable as you work across entries. The panel now:

• Stays open when you navigate between entries.
• Preserves its state during entry navigation within your session.
• Remembers a custom width and restores it when you open later entries.

This reduces the need to reopen or resize the Live Preview panel as you move between multiple entries.

Refer to the Live Preview documentation for details.

We’re adding enhancements to the Algolia app:

• Multiple Environment and Branch Support: Mapping rules can now be scoped to specific environments and branches. Multiple rules for the same environment are identified by different labels.
• Content Type Level Index Mapping: Index mapping is now configured one content type at a time, with an explicit index required for each rule.
• Enhanced Field Level Mapping: Field-level mapping now supports environment and branch-level filtering. Leaving the dropdowns empty defaults to all environments and all branches.

To learn more, refer to the Algolia App Installation Guide.

The Bulk Translate and Content Types Bulk Translate options in the XTM Full Page app now include a Source Language dropdown. You can now select any stack language as the source for translation, instead of being restricted to the stack language. By default, the master language is pre-selected.

To learn more, refer to the XTM App Installation Guide.

We are excited to announce Contentstack MCP Server v0.7.1, introducing management token authentication, explicit region selection, dedicated-infrastructure endpoint overrides, and a new guide for building custom MCP servers on top of Contentstack APIs.

What's New?

Management Token Authentication: The cma and cma-extended groups now accept either an OAuth session or a management token, making them suitable for headless and CI/CD environments where browser-based authentication is unavailable.

• Pass --management-token or set CONTENTSTACK_MANAGEMENT_TOKEN
• Must be used with an explicit region (--region or CONTENTSTACK_REGION)

Explicit Region Selection: Region can now be set independently of OAuth, so CDA-only and management-token-only setups route to the correct data center without an active OAuth session.

• Supported codes: NA, EU, AU, AZURE_NA, AZURE_EU, GCP_NA, GCP_EU
• Common aliases accepted (e.g., us, aws-na, azure_eu)
• The AU (Australia, AWS) region is now supported
• An unrecognized region value returns a hard error at startup

Dedicated-Infrastructure Endpoint Overrides: Customers on private or dedicated infrastructure can now override the base URL for any service individually. When set, an override fully replaces the region-derived base URL for that service.

Build Your Own MCP: Tool Definitions API: Developers can build a custom MCP server, agent framework, or API client on top of Contentstack APIs. It covers:

• Public tool-definition endpoints for all API groups
• Tool definition format and the mapper object
• Step-by-step request construction
• Base URLs, region resolution, and authentication
• Worked examples for CDA, CMA, and Launch GraphQL

To learn more, refer to the Contentstack MCP Server documentation.

Gain deeper visibility into your Contentstack activity with Log Targets within Administration, a powerful new feature that lets you seamlessly export system-generated logs to your own cloud storage. Whether you're monitoring activity, ensuring compliance, or integrating with observability tools, Log Targets gives you full control over your log data.

Key Features

• Flexible Log Exporting: Export key system logs, including audit, published, and webhook logs, directly to your preferred cloud storage for centralized monitoring and analysis.
• Multi-Cloud Support: Connect with leading cloud providers like AWS S3, Azure Blob Storage, and Google Cloud Storage to store and manage your logs securely.
• Automated Scheduling: Set up hourly export schedules to ensure your logs are consistently updated without manual intervention.
• Customizable Storage Structure: Define base paths within your storage buckets to organize logs efficiently by type or use case.
• Improved Visibility with History Tracking: Monitor export activity through a detailed history view, including status, duration, and error reporting for better troubleshooting.

With Log Targets, you can extend Contentstack logging capabilities beyond the platform and into your existing workflows, making observability, auditing, and analysis more streamlined than ever.

Check our documentation for full details, or contact support to get started immediately.

You can now roll back a deployed release in Contentstack to quickly recover from unintended deployments.  If a release is deployed to the wrong environment or publishes content before it is ready, create a rollback release or a duplicate of the deployed release, and deploy it to restore entries and assets to the version published before deployment.

Release Rollback provides a fast, predictable way to recover from publishing mistakes without manually restoring content. It reverts references and assets along with entries, and unpublishes documents that had no previous published version, so the environment returns to its last known good state in a few steps.

Refer to the Roll Back a Release documentation for more information.

We’ve added taxonomy-based filtering to the Select Entries modal used by Reference fields and embedded references in the Rich Text Editor fields. The modal now supports filtering using the existing Taxonomy column.

You can now:

• Filter entries by one or more taxonomy terms directly in the Select Entries modal.
• Find relevant entries faster in large content sets.

Note This brings the Select Entries modal in line with the taxonomy filtering already available on the Entry List page.

We are excited to announce Contentstack MCP Server v0.7.0, introducing a new CMA Extended tool group and comprehensive Contentstack MCP Server documentation.

What’s New?

CMA Extended: A new group of 22 tools that extends core CMA functionality with advanced administrative and governance operations.

• Audit log access and inspection
• Entry and asset version history
• Global field creation and updates
• Asset folder management
• Workflow and publish rule inspection

Contentstack MCP Server Documentation: A new guide is now available, covering:

• Setup, authentication, and configuration
• All API groups and tools
• Common workflows and best practices
• Troubleshooting and reference documentation

To learn more, refer to the Contentstack MCP Server documentation.

Gain a clearer, more actionable view of your organization’s security posture with the new Security Dashboard. Designed for admins and security-focused roles, this feature brings critical insights, activity tracking, and recommendations into a single, streamlined interface, helping you proactively manage risks and strengthen compliance.

Key Features

• Security Dashboard Overview: Access a dedicated dashboard that replaces basic organization info for elevated roles, providing a unified view of security metrics, risks, and activity.
• Current Trends Snapshot: Monitor key metrics like total users, inactive accounts, pending invitations, and users without MFA to quickly identify potential vulnerabilities.
• Security Scorecard: Understand your organization’s security posture with a dynamic score (Critical, At Risk, Secure) based on weighted controls like SSO, MFA, and session policies.
• Role Distribution Insights: Visualize how roles are assigned across your organization to prevent over-allocation of privileged access.
• Password Compliance Tracking: Identify users with outdated passwords through categorized timelines, enabling better enforcement of password policies.
• User Session Insights: Track active sessions, including long-lived sessions (60+ days), to reduce exposure from unattended logins.
• Recent Activity Feed: Audit security-related events such as user changes, role updates, MFA resets, and configuration changes with severity indicators.

Stay ahead of potential threats with smarter visibility and actionable insights. The Security Dashboard empowers your team to continuously strengthen security and confidently manage your organization at scale.

Read our documentation for more information.

Unified role-based access control (RBAC) introduces finer control over who can administer an organization and what they can do within each product, two new Administration roles, and the ability to create custom roles per product.

Key Features

• Redesigned user invite interface: Managing users and roles now happens in a completely redesigned interface that brings inviting users, assigning organization-level and product-level roles, and reviewing access into one consistent experience. The same experience extends to SSO identity provider role mapping, SCIM group-to-role mapping, and Teams.
• New Administration roles: Two out-of-the-box roles join Owner, Admin, and Member at the organization level:
  • Security Manager: Manages an organization's identity and security configuration, including SSO, SCIM provisioning, security settings, and webhooks, with read-only visibility into users, roles, and teams. This removes the previous dependency on the Owner role for configuring SSO.
  • Product Analytics Viewer: Previously, viewing Analytics required Owner or Admin role access. This new role provides read-only access to organization information and analytics, without the ability to change settings or manage users.
• Multi-role support: Users can now hold multiple organization-level roles at once. For example, a user can be both a Member and a Product Analytics Viewer without requiring elevated admin privileges.
• Granular product-level permissions: RBAC now scopes access within each product, not only across the organization. Administrators can assign different levels of access per product, such as Administration and Assets, so each user receives precisely the access required for their tasks.
• Custom roles per product: When the default roles do not match a team's responsibilities, administrators can create custom organization-level roles for a product in Administration by selecting permission categories and actions such as View, Create, Edit, and Delete.

What's Next for RBAC

This release is the first step in a broader rollout. Contentstack is extending RBAC adoption across the platform, with significant enhancements to CMS RBAC coming next, bringing the same granular, consistent access control to stacks and content management.

To learn more about Administration roles, product roles, and creating custom roles, refer to the documentation.

The Contentstack migration framework now supports Drupal as a legacy source and introduces single sign-on (SSO) authentication to the migration framework.

Key Enhancements:

• Drupal support: Migrate content from Drupal 8, 9, 10, and 11 into a Contentstack stack using the migration framework workflow. Connect your Drupal MySQL database, validate the connection, select your destination stack, and map your content types during migration.
• Asset migration from Drupal: Move images and media assets alongside your content by providing the base site URL and public asset path of your Drupal site during setup.
• Flexible content-type mapping: Use the structure detected from your Drupal data as-is, or map Drupal content types to existing content types in your destination stack when you've already modeled your content in Contentstack.
• Single sign-on (SSO) authentication: Authenticate the migration framework using OAuth 2.0. The tool requests only the permissions required to migrate content, keeping the access granted narrowly scoped to the task.

Note Drupal 7 is not supported. Drupal 7 has reached the end of life and is deprecated by the Drupal community.

For installation and setup, refer to the Migration Tool Setup Guide (Drupal) and the Authenticating the Migration Tool via SSO guide.

We are excited to announce the new Agent OS framework, featuring Agents and Polaris, and the launch of AI Credits. This update provides a more flexible and scalable way to manage and deploy artificial intelligence capabilities across your organization.

What’s New?

• Agent OS (Agents & Polaris): Step into the next generation of automation with Agent OS.
  • Agents: Deploy autonomous, task-oriented AI agents designed to handle complex workflows with minimal oversight.
  • Polaris: Meet your new platform-wide chat assistant. Polaris is integrated across the entire ecosystem to provide instant support, answer queries, and help you perform AI tasks across your content.
• AI Credits: We’re introducing a unified credit system to power your AI operations. This simplified consumption model provides greater visibility and control over your AI usage, ensuring your teams monitor and manage credits more effectively.
• Analytics: To help monitor usage and performance trends, dedicated analytics dashboards are now available for Agents, Polaris, and AI Credits.

This release transforms how organizations access, manage, and scale AI capabilities with centralized controls, intelligent automation, and enterprise-ready experiences.

To learn more, refer to the Agent OS and AI Credits documentation.

We've introduced two enhancements that help you manage entry relationships and URL structures more efficiently while working in Contentstack.

Configure Default URL Patterns for URL Fields

You can now configure default URL patterns for URL fields while creating or editing content types.

With URL pattern configuration, you can:

• Define consistent URL structures using static paths and dynamic placeholders
• Use supported system fields, custom text fields, locales, and taxonomies in URL patterns
• Automatically generate localized and structured entry URLs
• Apply formatting rules such as casing, word separators, JSON mappings, fallback values, and regex transformations
• Reduce manual URL creation errors and maintain consistent URL structures across entries

For example, you can define a pattern such as /blog/:locale/:field[title] to automatically generate localized blog URLs.

Refer to our documentation to learn how to configure URL patterns and formatting options.

Open Referenced Entries in a Sliding Panel

You can now open referenced entries in a sliding panel within the entry editor. Click the new Open in overlay icon next to a reference to view it without leaving the current page. You can also navigate deeper by opening up to five additional referenced entries in stacked panels and use breadcrumbs to switch between them.

This update reduces the need to open multiple browser tabs when working with nested references, making it easier to validate content relationships or make quick edits without losing your place.

Refer to the documentation on navigating referenced entries.

Note: This feature is plan-based and available to early access customers. Contact support to request access.

You can now publish entries in Contentstack with greater confidence and control with enhanced reference validation and preview workflows. The updated publishing experience helps you review, validate, and selectively publish referenced content before it enters the Publish Queue, reducing errors and rework.

What's New?

• Review referenced content before publishing
  Use the Publish Review modal to review the parent entry and all associated references to understand what is published before content enters the Publish Queue.
• Validate references before publishing
  Validate referenced entries and assets using the Validate Items modal to catch issues early and prevent publish failures.
• Filter and select references precisely
  Choose to publish only the parent entry, specific references, first-level references, or all references to gain greater control over what goes live.
• Monitor validation and publish progress
  Track validation tasks and publishing status through task windows to maintain visibility and confidence throughout the publishing process.

For more details, visit our documentation or contact our support team.

We are expanding the power of Automations by introducing real-time triggers for our most popular connectors. This update transforms your static workflows into reactive, event-driven systems that respond instantly to external actions.

New Connector Triggers

Our existing connectors for Cloudinary and Netlify have been updated with new trigger options to help you automate your media, deployment, and communication pipelines.

• Cloudinary: Automate your digital asset management by triggering workflows based on any changes within your Cloudinary environment. Supported events include asset uploads, metadata modifications, folder creation, access control shifts, etc. Learn more about Cloudinary Trigger.
• Netlify: Keep your development and content teams in sync with your deployment pipeline. You can now trigger automations when builds start, deployments are created, or when environments are locked or unlocked due to permission or configuration issues. Learn more about Netlify Trigger.

Navigation Updates

To support the continued expansion of our platform, we’ve streamlined the project navigation for a more scalable and organized experience.

Settings-related options are now consolidated within the Settings panel in the top navigation, making configuration and operational controls easier to access and manage.

You can now find the following under Settings:

• Connected Apps
• Audit Log
• Execution Log
• Project Variables

Platform Discovery is your new dashboard for exploring everything Contentstack has to offer. Get a unified view of every capability available to your organization, and spot the ones you are missing out on.

With this release you can:

• Audit your stack: See a complete inventory of your features in one glance.
• Find hidden value: Instantly identify features marked No Recent Activity, powerful tools you already have access to but aren’t leveraging.
• Scale intelligently: See clear pathways to advanced capabilities (like Personalization or AI) that align with your growth goals.

Where to find it: Open the "App Switcher" in the top navigation bar and select Platform Discovery.

Read our documentation for more information.

We’ve introduced new taxonomy enhancements in Contentstack that improve how teams manage, publish, release, and govern taxonomy, making it easier to manage taxonomy structures across environments and locales.

Publish Taxonomies Independently

You can now publish taxonomies independently, similar to entries. This allows you to manage taxonomy structures separately from content publishing workflows.

With Taxonomy Publishing, you can:

• Publish an entire taxonomy and its complete term hierarchy
• Publish to specific environments and locales
• Select a branch and apply locale fallback hierarchies
• Schedule publishing jobs (if enabled)
• Automate taxonomy publishing through APIs

Learn more about publishing a taxonomy.

Nested Reference Publishing Now Supports Taxonomies

Nested Reference Publishing now includes taxonomy support. When you publish an entry using Send with References, any referenced taxonomies:

• Appear in the Publish References modal
• Are included in the publish job
• Publish their complete hierarchy to the selected environments and locales

This helps maintain consistency between published entries and their associated taxonomy structures across environments.

Learn more about working with Nested Reference Publishing.

Add Taxonomies to Releases

You can now add taxonomies directly to releases and deploy taxonomy structures alongside related content changes.

You can add taxonomies to a release from:

• The Taxonomy Details page
• The Taxonomy List page
• Entry references, where associated taxonomy terms are automatically included when references are added

This helps teams schedule taxonomy deployments, streamline localization workflows, and reduce publishing inconsistencies.

Learn more about adding a taxonomy to a release.

Granular Taxonomy Permissions for Custom Roles (Early Access)

You can now assign taxonomy management permissions to custom roles without granting full administrative access to the stack.

Custom role users can manage taxonomy structures using permissions such as:

• Create
• Read
• Update
• Publish/Unpublish
• Delete

You can also apply exceptions to restrict access to specific taxonomies or actions.

This enhancement separates taxonomy structure management from entry-level taxonomy permissions, giving teams more precise access control and governance flexibility.

To learn more, refer to the Permissions on Taxonomies documentation.

We’re excited to introduce Contentstack Assets, a modern, AI-powered digital asset management experience built for teams that need to organize, govern, and reuse assets at scale.

This upgrade moves beyond stack-based asset management to deliver a centralized, structured, and reusable asset system across your organization. With support for shared asset spaces, custom metadata modeling, localization, AI-powered enrichment, and advanced governance controls, Contentstack Assets helps teams streamline operations and deliver consistent digital experiences faster.

Key Features

• Centralized Asset Management Across Stacks: Manage assets in dedicated spaces and reuse them seamlessly across multiple CMS stacks, reducing duplication and improving consistency.
• Spaces and Workspaces for Better Organization: Organize assets by brand, team, campaign, or region using independent spaces and isolated workspaces for experimentation and campaign workflows.
• Custom Asset Modeling: Create custom asset types and user-defined metadata fields to structure assets around your business needs, from campaign information to licensing and compliance data.
• Native Asset Localization: Manage multilingual versions of assets within a unified workflow, including localized metadata and asset variants with fallback language support.
• AI-Powered Asset Enrichment: Automatically generate alt text and tags for images to improve discoverability, accessibility, and SEO while reducing manual effort.
• Visual Markups for Interactive Experiences: Add hotspots and bounding boxes to images to power shoppable content and richer visual experiences.
• Advanced Search and Filtering: Find assets faster using metadata-driven filters, saved views, color-based filtering, and intelligent search capabilities.
• Asset Versioning and Governance: Track changes to both metadata and binaries with built-in version history and rollback support.
• Enterprise-Ready Access Control: Manage permissions using predefined and custom roles with support for SSO and SCIM-based user provisioning.

Contentstack Assets lays the foundation for a more intelligent, scalable, and future-ready asset strategy. We’re excited to see how your teams use it to create faster workflows, stronger governance, and richer digital experiences.

Check our documentation for full details, or contact support to get started immediately.

We are excited to introduce Clearwinner, a new Marketplace app designed to automate the post-test cleanup process for finished A/B test experiences.

Traditionally, concluding an A/B test requires manually identifying variant changes, merging them into baseline entries, publishing updates, and archiving the test. Clearwinner eliminates this time-consuming workflow with a single-click solution.

Key Features:

• Automated Winner Identification: Easily view tests where Personalize has identified a winning variant based on statistical confidence levels (Has_Won, Leading_significantly, or Leading).
• One-Click Merge & Publish: Automatically merge winning variant content into your baseline entries and publish the updates to your live environment in one background process.
• Integrated Cleanup: Automatically archive concluded tests in Personalize and permanently delete redundant variant data to maintain a clean CMS stack.
• Bulk Operations: Select and process multiple winning tests simultaneously to streamline large-scale experimentation cycles.

To learn more, refer to the Clearwinner App Installation Guide.

We’re improving the Mappings feature in Advanced Settings to make it clearer how values are resolved and how to work with nested configuration data. It now includes a dedicated Template Substitution section and better examples for nested values.

Dynamic values can already be used in request headers and body, and now they also work in URLs and query strings.

With a single syntax, {{ var.NAME }}, you can apply dynamic values across the entire outbound request. This enables API versioning, multi-tenant routing, pagination, and more, all driven from your configuration without hardcoding.

Learn more about Introduction to Advanced Settings in Developer Hub.

Here’s an overview of the latest updates in Contentstack Launch:

• User-Agent Support in Cache Priming:

  Cache priming requests now include a dedicated User-Agent header, allowing you to easily identify and filter them from regular traffic in your logs.

  Learn more in Identifying Cache Priming Traffic.

• Cancel Active Deployments:

  You can now cancel an ongoing deployment in Launch, allowing you to stop unintended updates before they are applied.

  Learn more in Cancel Deployment.

• Support for Node.js 24.x:

  We've added support for the Node.js 24.x runtime, ensuring compatibility with the latest Node.js features and improvements.

  Refer to the Supported Node.js Versions.

• Build Machines for Launch Builds:

  Launch now offers multiple build machine tiers, enabling you to allocate appropriate CPU, memory, and disk resources for faster and more reliable builds.

  To learn more, refer to Build Machines on Launch.

• Server Machine Tiers for Launch Functions:

  With new server machine tiers in Launch, you can control CPU and memory allocation for your functions, ensuring better performance and scalability based on workload needs.

  For more details, refer to Server Machines on Launch.

• Launch “Warming Up” Screen for Idle Environments:

  Launch now supports an automated warm-up state for inactive environments, where sites briefly scale down and quickly reinitialize on the next request to ensure efficient resource usage without impacting availability.

Managing entries just got faster. You can now switch between Publish Status, Release Status, and Publish Rules directly from the Entry Status panel using a dropdown. You can also apply filters to refine the displayed data. Selected filters remain applied across all views.

This update helps you focus on specific entry details without navigating multiple sections. For example, you can filter by environment and quickly switch views to check publishing status, release progress, or approval rules in the same context.

Refer to the view entry status documentation for more details.

We’re adding enhancements to the Algolia app:

• Stack Delivery Token Support: Added a new section in app configuration to fetch published data via the Content Delivery API for improved data synchronization.
• Bulk Field Mapping: You can now add up to 30 field paths at once in the Mapping section by pasting values separated by commas, spaces, or semicolons, or by pressing the enter key.
• Enhanced Search: Introduced a search function to find entries by Entry UID and Title in the Algolia Full Page app.

To learn more, refer to the Algolia App Installation Guide.

As part of our ongoing efforts to enhance platform security and reliability, we are deprecating legacy Content Management API versions /v1 and /v2. This change ensures that all users benefit from the latest performance improvements, security standards, and feature capabilities available in /v3 APIs.

ImportantThe legacy /v1 and /v2 CMA endpoints are now deprecated and will be officially decommissioned in 60 days.

We recommend reviewing your integrations and updating them to use the current API version to avoid any disruption.

• Deprecation of /v1 and /v2 APIs: Legacy API versions will be phased out and will no longer be supported.
• Planned Access Restrictions:
  • /v1 endpoints will be blocked across all cloud environments.
  • /v2 endpoints will be restricted to AWS North America (AWS-NA) only during the transition period.
• Shift to /v3 APIs: The latest API version is now the standard and actively supported across all environments.

This update helps us deliver a more secure and consistent API experience. We’re committed to supporting your transition, upgrade to /v3 today and take advantage of the latest improvements.

We’ve added enhancements to the XTM app:

• Unified Field Management: Replaced separate inclusions or exclusions settings with a single Manage Fields section for easier control.
• Multi-Branch Support: You can now translate entries in non-main branches.

  Note Existing users must update the app to activate this feature.

• Unlimited Bulk Translation: Removed the entry translation limit for faster, high-volume processing.
• Global Exclusions: Added a dropdown to exclude specific fields across all content types.
• Automatic Filtering: Non-localizable fields are now automatically excluded from translation by default.

To learn more, refer to the XTM App Installation Guide.

Managing user access just got more powerful. Admins and organization owners can now take immediate action to secure their organization by logging out multiple users at once.

Select users in bulk from Administration > Users and immediately terminate their active sessions. Once triggered, all active sessions for the selected users are invalidated, ensuring they are logged out across all devices without delay. Take control of your organization’s security with faster, more efficient session management.

Check our documentation for full details, or contact support to get started immediately.

We’ve enhanced the Bulk Operations app to support larger datasets and smoother workflows:

• Release Version 2.0: Supports references, higher selection limits, and improved API behavior. The release version available for the stack depends on your organization plan.
• Revamped Tables with Pagination: All modules (Entries, Assets, Releases, and Find and Replace) now use a paginated table for improved performance and usability, along with upgraded item limits.
• Configurable Asset Upload Limits: You can now upload assets based on your organization’s configured upload limit to better manage bulk assets operations.

To learn more, refer to the Bulk Operations App Installation Guide.

You can now use English - Uzbekistan (en-uz) as a locale in Contentstack to manage English-language content scoped to the Uzbekistan region.

Use this locale to manage region-specific variations, such as pricing, date formats, currency and localized messaging, without duplicating content or approximating wi.

To get started, add English - Uzbekistan (en-uz) from your stack’s Languages settings and create localized entries.