Version 2.11.0+
These are the release notes for the v2.11.0 release of Witboost.
Please refer to the official documentation for a more in-depth overview of the released features.
Features
Control Plane
Approval gates for system deployment
Introduced Deploy Approval, a control plane feature that adds an approval gate before deploy requests can be executed. When active, data teams must submit a deploy request that waits for a designated approver to review and either approve or reject it, rather than deploying directly.
The activation scope is fully configurable: platform teams can restrict the gate to specific system types (e.g., only Data Products) and environments (e.g., only production), or apply it broadly across all systems and environments.
Platform teams can enable and customize this feature from the Workflows section of the Administration Panel; it is disabled by default.
Check the documentation for additional information.
Lifecycle Manager Team Role
The Lifecycle Manager has been added as new Team Role, where users can now be assigned to this role, which can be configured as target of workflow engine notification and requests, allowing the project owner to delegate data project lifecycle management, deployment approvals and more to other users.
Check the documentation for additional information.
Dynamic Data Enrichment for Descriptors
Project descriptors can now be enriched with live platform state at evaluation time through Dynamic Descriptor Enrichment Providers, enabling governance policies and Tech Adapters to make decisions based on current conditions rather than static, release-time data. Providers are registered automatically and can be independently enabled, configured, and scoped to specific project types from the Descriptor Enrichment Providers section of the Administration Panel.
The first available provider is the Deployment Status Provider, which adds per-environment deployment status to the descriptor. All providers are disabled by default. Before enabling a provider, especially in a production environment, carefully review the documentation on how to configure providers and follow the rollout guide to verify that your existing Tech Adapters and governance policies can handle the additional fields in the enriched descriptor, avoiding unexpected provisioning failures.
Check the Dynamic Data Providers overview and the Deployment Status Provider documentation for additional information.
Custom View for descriptor preview in Edit & Test
A new set of custom views have been introduced to show a visual representation of the descriptor for a project. These views are rendered alongside the YAML descriptor on the Edit and Test tab of the project page in the builder, taking the rendered descriptor as input, including any environment specific information if present and useful to show the project's metadata to be deployed in a more user-friendly way. You can enable this feature from the Configuration section of the Administration Panel as it's disabled by default.
Check the Custom Views documentation for additional information.
Code block visualization on Templates and Custom Views
You can now show code-formatted text as part of Custom Views and Templates. A new component code_block has been introduced for Custom Views that allows showing a string or object field as formatted code with syntax highlighting. On Templates, string fields now support format: codeblock which renders the field as an interactive code editor with syntax highlighting for most languages.
Check the documentation for additional information.
Skeleton Entity refs and RDF parsing helper
Skeleton Entities now support refs, allowing reusable values to be resolved during descriptor rendering. Refs can load files from the current entity repository, such as schemas, cluster settings, or other descriptor inputs, and can also read selected values from descriptors published in the Marketplace.
This enables richer descriptor generation for use cases such as including repository files in the generated descriptor, reusing parsed YAML, JSON, or text values, and deriving fields from upstream published descriptors. A new parseRdf Nunjucks function is also available to parse RDF content and inspect semantic metadata in formats such as Turtle, RDF/XML, N-Triples, N-Quads, TriG, N3, and JSON-LD.
Check the refs documentation and Nunjucks helpers documentation for additional information.
Business Concept support in Entity Search Picker
The Entity Search Picker now supports Business Concepts, allowing users to select one or more concepts from the Witboost Knowledge Graph Manager (KGM) when filling forms. This makes it possible to associate descriptors with glossary terms, concept schemes, or other business terminology managed through KGM.
Check the documentation for additional information.
Governance
Managed Policies and Environment Promotion Strategy
Introduced Managed Policies, a new category of governance entity whose validation logic is built into Witboost's Computational Governance. Unlike custom policies — where you write CUE, natural language, or host a remote service — Managed Policies are configured through a guided wizard and evaluated natively by the platform, enabling governance patterns that depend on data only the platform itself owns.
The first available Managed Policy type is the Environment Promotion Strategy, which enforces deployment ordering across environments. A single policy expresses the entire promotion chain (e.g., DEV → QA → PROD), ensuring that a system must be deployed to one or more prerequisite environments before it is allowed to be deployed to a target environment.
Check the Managed Policies overview and the Environment Promotion Strategy documentation for additional information.
Market Plane
Business Concept Map support for Consumables
The Business Concept Map can now show not only Products, but also linked Consumables associated with a selected Business Concept. Business Concepts assigned to project components are reflected in the Marketplace, helping users discover the specific consumables that match a business term.
Access Control pending requests view
Each product's Access Control panel now includes a new Pending Requests view, giving resource owners (and Data Access Managers) a dedicated place — outside the Notification Center — to review, sort, filter, approve or reject pending access requests directed at that product's consumables, individually or in bulk.
Check the documentation for additional information.
AI Capabilities
Witty Skills
Introduced Skills, modular prompt sections that define Witty's behavior, persona and grounding rules. Platform teams can add, edit, enable/disable and delete skills directly from the Administration Panel (Extension Manager > Witty Skills) without redeploying. Witty ships with a set of default skills, and changes take effect immediately.
Check the documentation for additional information.
Witty Chat History
Witty now persists conversation history across sessions, allowing users to resume previous chats and review past interactions. Chat history is accessible directly from the Witty chat panel.
Check the documentation for additional information.
Witty Streaming Responses
Witty now streams responses in real time as they are generated by the underlying LLM, providing immediate visual feedback and a more responsive conversational experience instead of waiting for the full response to be produced before displaying it.
Check the documentation for additional information.
Platform & Installation
Git Providers Configuration in the Administration Panel
Git Providers can now be configured directly from the Administration Panel, allowing platform teams to add, edit, enable/disable and delete Git Providers without redeploying. The configuration is stored in the platform's database and takes effect immediately.
Check the documentation for additional information.
RBAC role cloning
Administrators can now clone an existing RBAC role from Administration > Roles & Permissions, creating a new role with the same permissions as the original one. During the clone flow, administrators can also choose whether to copy the existing role assignments, making it faster to create similar roles for different teams, domains, or operating models.
Data Landscape filter in Practice Shaper graph
The Practice Shaper graph in Administration > Practice Shaper > Graph now includes a Data Landscape filter, allowing administrators to focus the graph visualization on a specific Data Landscape for clearer navigation and analysis.
Pod labels, PDB, and scheduling support for all service charts
All Witboost Helm subcharts now support pod annotations, pod labels, node affinity, tolerations, and Pod Disruption Budgets (PDB). This enables operators to meet cluster policies — such as Gatekeeper or Kyverno rules — and to control pod placement and disruption behaviour across all Witboost services.
Hook jobs migrated to dedicated witboost-hooks chart
The pre-install/pre-upgrade hook jobs have been extracted from the Witboost umbrella chart into a dedicated subchart (witboost-hooks). The new hook uses a purpose-built Witboost image instead of curlimages/curl, with hardened security context defaults. The Infra CLI handles the migration automatically; see the migration guide for details.
HTTPRoute timeout support
Gateway HTTPRoutes have been extended to allow specifying timeouts under route rules in Witboost subcharts, giving operators control over request and backend timeout settings.
Witboost App CLI
Build Descriptor --include option
The build-descriptor command now includes a repeatable --include option for adding local files, folders, or glob patterns to the generation input. This allows Skeleton Entity refs to resolve the referenced files when rendering descriptors.
Issues fixed and minor improvements
v2.11.0
- Accessibility Improvements
- Security Enhancements
- Improved Infra CLI
prepare-secretcommand with better defaults, input hints, and a dedicated license-update option - Infra CLI
mirror-imagecommand now includes alocalpassthroughoption for environments where direct registry-to-registry communication is not available - Improved Infra CLI error handling for
mirror-image,mirror-chart, andcheckdbcommands when the CLI has not been initialized - Fixed
caCrtFilePathdeprecation notice in CustomCA configuration for Coordinator, KGM, WCG, and UI - Fixed
remoteSecretNameandtlsCertSecretNamemutual exclusivity validation across Coordinator, KGM, and Governance charts - Updated mTLS documentation for Coordinator, CGP, UI, and KGM
- Updated Infra CLI brew tap installation instructions
v2.11.1
- Added workflow management capabilities for workflows, including versions, management UI, deploy approval enhancements, and customizable approval notifications
- Fixed a bug where deleting a service account would fail, as well as when creating a new service account or creating an access token for one would fail under certain ingress configurations
- Fixed an issue where Custom URL Picker fields using dynamic parameters could appear empty when reopening edit forms, even though the saved value was still present. Previously saved values are now preserved correctly during form loading
- Improved GitLab repository provisioning resilience during publish step in creation templates. When GitLab temporarily rejects a newly created project request because subgroup or permission changes have not fully propagated yet, Witboost now retries the operation with backoff before failing
- Fixed an issue where dependent fields were not populated when opening templates with a prefilled EntitySearchPicker