Daily Digest

Daily Digest – 8 June 2026

2026 Capabilities Progress

  1. Page generation — Active development continues on Canvas AI, with work addressing code generation reliability issues including unsupported library usage [86], prop/slot name collisions [87], and component editing problems [85]. A client-side orchestration loop proof-of-concept is underway [63], representing foundational infrastructure for AI-driven page assembly.
  2. Context management — The Context Control Center is nearing release candidate with performance optimizations in progress [127][135], cost reduction testing underway [124], and improved display interfaces being developed [123]. Work on conditional subcontext providers [130] and stable selection pipeline events [134] will enable more sophisticated context orchestration.
  3. Background agents — Progress on queue processing efficiency, with a critical issue identified where only one automator worker queue item processes per cron execution [9]. The agent configuration UI is under discussion to reduce complexity [5], which affects administrator ability to set up and manage autonomous agents effectively.
  4. Design system integration — Canvas saw significant stability improvements with fixes for component hydration failures [46], config schema validation errors [47], and prop shape changes [79][80]. Work on component auditing tools [57][107] and improved CLI configuration [61][106] supports better design system management and validation workflows.
  5. Content creation and discovery — Integration work progressed on AI Search to restore 2.x branch compatibility [144], though this represents maintenance rather than new capability. The CKEditor streaming content generation issue with Gemini 2.x models remains unassigned [1], blocking content drafting for users of that AI provider. Vector search capabilities for ECA remain in planning [32].
  6. Advanced governance — Canvas made substantial progress with conflict detection infrastructure under active development [71][108][110][113][122], translation deletion routes completed [52][98], and language-based workflow controls implemented [70][90]. Work continues on symmetric translation propagation [49][92][93] to ensure governance rules apply consistently across language variants.
  7. Intelligent website improvements — No direct progress today.
  8. Multi-channel campaigns — No direct progress today.

Shipped

  1. Canvas translation workflow controls — The system now prevents editors from creating Canvas pages in non-default languages [70][90], enforcing proper source-then-translate workflows and reducing content management errors.
  2. Canvas translation deletion capabilities — New routes enable deletion of translations across both Canvas Pages and Content Templates [52][98], completing essential translation management workflows.
  3. Canvas language switcher expansion — Language switching support now extends to Content Templates [78][119], allowing editors to manage translations across all Canvas content types.
  4. Canvas CLI configuration improvements — Configuration settings moved to canvas.config.json with more intuitive default paths [61][73][106][117], reducing setup friction for development teams.
  5. Canvas autosave fix for multilingual sites — Resolved 409 conflict errors on non-English monolingual sites [59], eliminating workflow disruptions for international deployments.

Ongoing

  1. AI Agents interface consistency updates — Three merge requests addressing spelling errors, menu labeling, and navigation structure are awaiting review [35][37][38][39][41][42], needed for professional administrator experience.
  2. Canvas entity reference selection backend — Backend APIs for entity reference selection are under development and awaiting security review before stable release [65][102], currently blocking production readiness.
  3. Canvas symmetric translation behavior — Work continues on ensuring component updates propagate correctly across all language translations [49][93], critical for maintaining content consistency in multilingual deployments.
  4. Canvas conflict detection during autosave — Development is active on detecting and handling editing conflicts during autosave operations [71][113], preventing data loss when multiple editors work simultaneously.
  5. Context Control Center cost optimization — Teams are testing cheaper AI models to reduce operational costs from current $1-2 per page [124], essential for broader market adoption.

AI (Artificial Intelligence)

Activity over the past 24 hours focused on issue assignment and planning rather than code delivery. The team continued to organize work across several key areas: improving compatibility with AI providers, refining user experience, and strengthening the module's architecture.

Two critical bugs received attention. A streaming compatibility issue with Gemini 2.x models that causes failures when generating content was assigned to Joshua Fernandes (joshua1234511). Separately, work on guardrail plugins that only evaluate the last message in conversations remains unassigned—this affects safeguards designed to filter inappropriate content across full chat histories.

Several architectural improvements are in progress with merge requests already open. These include allowing vector database providers to declare themselves as default and modernizing provider configuration forms. A code quality refactoring to improve maintainability was also assigned to Joshua Fernandes.

The team is also addressing navigation and usability concerns, including reorganizing menu structure to improve discoverability and ongoing discussions about simplifying the agent configuration interface.

How can I help on this project?

  • Assign a developer to the conversation-wide guardrail evaluation issue, which affects content safety features but currently has no owner.
  • Consider UX resourcing for the agent configuration interface discussion, which would benefit from design expertise to reduce complexity.

View project on GitLab: ai

Drupal AI Initiative

Activity over the past 24 hours focused on organizational operations as the project welcomed a new Gold partner and streamlined its administrative processes.

Kristen Pol (kepol) is leading onboarding for Mearra, the initiative's newest Gold partner. The Onboard Gold AI partner: Mearra work includes integrating them into communication channels, team meetings, and project resources. Related tasks include updating the public-facing partner pages and preparing an audit to verify successful onboarding. This partnership expands the commercial ecosystem supporting Drupal's AI capabilities.

Several administrative improvements were completed. The project closed multiple template-related issues as part of consolidating workflow tools. Work on Convert partner templates into GitLab templates aims to standardize how new partners are onboarded, reducing administrative overhead for future partnerships.

Marketing work also progressed, with several content and planning issues being migrated to a dedicated marketing project space for better organization. This includes closing items related to the marketing sprint, messaging guides, and capability showcase articles that communicate the project's value to potential users and partners.

How can I help on this project?

  • Facilitate introductions between Mearra and other strategic partners or customers to accelerate partnership value realization.
  • Review and approve resource allocation for the marketing function to ensure content production keeps pace with partnership growth.

View project on GitLab: ai_initiative

AI Integration - ECA

This module extends Drupal's automation capabilities by connecting AI services with the ECA (Event-Condition-Action) framework, allowing site builders to create workflows that leverage artificial intelligence without writing code.

Activity over the past 24 hours was limited to issue tracking. A test issue created by Jürgen Haas (jurgenhaas) for notification testing was opened and closed. More significantly, Marcus Johansson (Marcus_Johansson) opened a feature request to add an ECA action for vector search. This proposal addresses a current gap: users can generate AI embeddings (numerical representations of text) through the module, but cannot yet use those embeddings to search a vector database and retrieve similar content. Adding this capability would complete the workflow, enabling site builders to create automated processes that find and act on semantically similar content—useful for recommendations, content discovery, and knowledge base applications.

No code was merged or committed during this period. The vector search feature remains in the planning stage with no assigned developer.

How can I help on this project?

  • Allocate developer resources to implement the vector search feature, which would unlock significant new automation capabilities for content discovery and recommendations.
  • Clarify strategic priority for AI-powered search features to help the team prioritize this request against other module enhancements.

View project on GitLab: ai_integration_eca

ai_provider_amazeeio

A new bug was reported during the period affecting administrators using configuration management tools. The issue array_flip(): Can only flip string and integer values in config_ignore_ignored_alter() causes PHP warnings to appear when attempting to import, export, or check the status of site configuration. This occurs when the module tries to load security keys without first checking whether the values exist, disrupting routine configuration management workflows.

The issue has been identified and documented by Ron Raney (rraney) and is currently awaiting review. No code changes were merged or committed during this period, so the bug remains unresolved. For a module in active development (version 1.3.x-dev), open bugs awaiting review are normal, though this particular issue may affect day-to-day site administration tasks for teams using the module.

How can I help on this project?

  • Allocate developer time to review and fix the configuration management bug, as it impacts routine administrative workflows.
  • Clarify project resourcing priorities if multiple issues are competing for attention.

View project on GitLab: ai_provider_amazeeio

AI Agents

The AI Agents module saw focused activity on user interface improvements and documentation clarity over the past 24 hours. Contributors addressed several interface inconsistencies and labeling issues that affect how administrators navigate and configure the system.

Three fixes related to navigation and terminology are progressing through review. Work is underway to correct a spelling error in configuration help text, update the admin menu label for consistency, and align the module's menu structure with AI 1.4.x categorization standards. These changes ensure administrators see consistent, professional labeling throughout the system. Prabhavathi Vanipenta (prabha1997) and Michael Anello (ultimike) contributed to these improvements.

A technical configuration issue is also being addressed: Sven Decabooter (svendecabooter) submitted a solution for a schema validation error affecting the CKEditor integration, which currently triggers system warnings.

Looking ahead, Marcus Johansson (Marcus_Johansson) is exploring improvements to how administrators restrict and control tool behavior, particularly for complex configuration scenarios.

How can I help on this project?

  • Prioritize review and approval of the three interface consistency fixes currently awaiting review, enabling a more polished administrator experience.
  • Provide user experience input on the tool restriction system discussion to guide strategic direction before development begins.

View project on GitLab: ai_agents

Tool API

The Tool API project saw development activity focused on improving compatibility with the AI module and advancing file handling capabilities.

Marcus Johansson submitted a solution to address a compatibility issue with the AI module through Add contextdefinitionoverrides to the ToolPluginBase. This work ensures that when the Tool API is used alongside the AI module, each instance maintains its own unique context definitions. The fix prevents conflicts that were occurring across all supported versions of the AI module. A merge request is ready for review.

The team also continued work on File Input Handling, which addresses how AI agents can securely work with files through the Tool API. This research and proof-of-concept effort will inform decisions about whether file handling features should be built into this module or developed separately, with a focus on secure input formats and file processing.

How can I help on this project?

  • Assign technical reviewers to evaluate the pending compatibility fix, enabling faster integration with the AI module.
  • Clarify strategic priorities around file handling to help the team decide whether to build capabilities in-house or develop a separate module.

View project on GitLab: tool

Drupal Canvas

Over the past 24 hours, the Canvas team completed significant work on translation capabilities and critical bug fixes ahead of stable release.

The team delivered translation infrastructure improvements including routes for deleting translations across both Canvas Pages and Content Templates, and extended the language switcher to Content Templates. Work also merged preventing Canvas pages from being created in non-default languages, ensuring editors follow proper source-then-translate workflows. Several critical bugs were resolved: autosave-triggered 409 conflict errors on non-English sites and an issue where the editor broke when accessed via language-prefixed URLs.

The CLI tool saw usability improvements with configuration moving to canvas.config.json files and more intuitive default paths. Test stability received attention through fixes to date-time widget tests and Playwright coverage improvements.

Active development continues on symmetric translation behavior (ensuring component updates propagate correctly across languages), conflict detection during autosave operations, and entity reference selection interfaces. Contributors Ben Mullins (bnjmnm), Ignacio Sánchez Holgueras (isholgueras), Kunal Sachdev (kunal.sachdev), Bálint Kléri (balintbrews), Dave Long (longwave), and Feliksas Mazeikis (f.mazeikis) drove the completed work.

Two critical-priority issues remain under active development: config schema validation errors affecting fallback components and refinements to symmetric translation handling for content entities.

How can I help on this project?

  • Support security review resourcing for the entity reference selection backend APIs, currently marked as a blocker requiring security sign-off before stable release.
  • Help prioritize decisions on theme dependency constraints, which currently prevent theme changes once Canvas content exists—this affects site maintenance flexibility.

View project on GitLab: canvas

Context Control Center (CCC)

The project is nearing its first release candidate, with teams focused on performance optimization, user experience improvements, and integration testing. Activity over the past 24 hours centered on stabilizing core functionality and preparing for broader adoption.

Work continued on making the system more cost-effective for users. Teams are testing with cheaper AI models to reduce operational costs, which currently run $1-2 per page with premium models. This could significantly lower the barrier to adoption.

Several performance and quality improvements are in active development. Contributors are addressing code consistency issues and optimizing how the system handles complex queries to reduce latency and processing costs. Work is also underway on redesigning the context item display pages to make information easier to scan and navigate.

The team closed one administrative issue related to publishing documentation, improving accessibility to implementation guidance. A quality issue was also resolved to ensure code validation runs correctly.

Integration exploration began with two external projects that are building on CCC's capabilities, which will help validate the architecture and inform future development priorities.

How can I help on this project?

  • Prioritize resourcing for model cost testing to determine if budget-friendly configurations can meet quality requirements before the release candidate ships.
  • Facilitate stakeholder review of the new context display design to ensure it meets end-user needs before implementation is finalized.

View project on GitLab: ai_context

References

[1] ai_ckeditor: streamGenerateContent causes 400 Bad Request with Gemini 2.x models (ai, Issue)

[5] (Discuss) Improve the add/edit Agent UI to make it less overwhelming (ai, Issue)

[9] Only one AI Automator worker queue item is processed per cron execution (ai, Issue)

[32] Add an ECA action for vector search (embeddings input → results of id, label, score) (ai_integration_eca, Issue)

[35] Spelling fix: "You will have to provider a JSON schema of the output wanted." (ai_agents, Issue)

[37] Change "AI Agent" to "AI Agents" in admin menu (ai_agents, Issue)

[38] AI Agents admin menu does not follow AI 1.4.x menu categorization (Tools & Automation) (ai_agents, Issue)

[39] Fix spelling - provider to provide in structured output help text (ai_agents, MR)

[41] Change 'AI Agent' to 'AI Agents' in admin menu label (ai_agents, MR)

[42] Add ai_agents.links.menu.yml to place AI Agents under Tools & Automation (ai_agents, MR)

[46] Fix AssertionError crash during component tree reconstruction when a component fails hydration (canvas, Issue)

[47] Fix LogicException on component instance updates and config schema validation errors on fallback components (canvas, Issue)

[49] Symmetric translations for content entities should allow only translatable properties in inputs to be overridden (canvas, Issue)

[52] Create Routes for Deleting Translations (canvas, Issue)

[57] Provide a Drush command to audit components (canvas, Issue)

[59] 1.5.0: Autosave triggers repeated 409 conflict errors on non-English monolingual sites (canvas, Issue)

[61] Move CLI sync settings to canvas.config.json, enable everything by default (canvas, Issue)

[63] Canvas AI: Client-side orchestration loop for Canvas AI(POC) (canvas, Issue)

[65] Backend APIs for Entity Reference Selection (canvas, Issue)

[70] Prevent creation of Canvas pages in a non-default language (canvas, Issue)

[71] Detect conflicts for Page entities during the handling of auto-saves/pending calls (canvas, Issue)

[73] Change default global CSS path in CLI and Workbench config (canvas, Issue)

[78] Extend language switcher support to Content Templates (canvas, Issue)

[79] Introduce a JsComponentInstanceUpdater that allows prop shape ("type") changes and accepts data loss (canvas, Issue)

[80] Coerce prop values across compatible types on JS component prop-shape change (instead of dropping) (canvas, Issue)

[85] Canvas AI: Unable to edit code components (canvas, Issue)

[86] AI continues to use unsupported React libraries while creating a component (canvas, Issue)

[87] AI can generate invalid code components: props' names may reuse slots' names (canvas, Issue)

[90] #3591625 "Prevent creation of Canvas pages in a non-default language" (canvas, MR)

[92] Draft: feat(Component sources): #3591596 Symmetric translation: component instance version updates must propagate to all translations" (canvas, MR)

[93] Resolve #3583684 "Symmetric content inputs" (canvas, MR)

[98] Issue #3591588: Use new DELETE route for language switcher. (canvas, MR)

[102] Draft: feat: #3585354 add content-entity-types / fields endpoints for entity reference selection UI (canvas, MR)

[106] feat(CLI Tool): #3591633 Move CLI sync settings to canvas.config.json, enable everything by default (canvas, MR)

[107] feat(Drush): #3585531 Add commands to list, audit, and generate Canvas components (canvas, MR)

[108] Draft: fix(Conflict detection): Exclude computed internal fields from Page entity normalization (canvas, MR)

[110] Draft: Detect conflicts during publishing [BE] (canvas, MR)

[113] Draft: feat(Conflict detection): #3589076 "Detect conflicts during auto-save polling" (canvas, MR)

[117] chore(CLI Tool): #3591632 Change default global CSS path in CLI and Workbench config (canvas, MR)

[119] #3590577 Language preview for templates (canvas, MR)

[122] Add conflict detection and resolution UI for selective publishing (canvas, MR)

[123] Add full view layout for context item canonical and revision pages based on new design (ai_context, Issue)

[124] Test Chicago Driesnote repo with cheaper models (ai_context, Issue)

[127] Clean up CCC inconsistencies and DX issues before rc1 (part 2) (ai_context, Issue)

[130] Conditional Subcontext Provider Calls (ai_context, Issue)

[134] Add stable selection pipeline events for contrib extension (ai_context, Issue)

[135] CCC Usage Tracking Query/Save Pattern (ai_context, Issue)

[144] Issue #3584015: Fix 2.x branch integration with AI 2.x (ai_search, MR)

Period: 8 June 2026 to 9 June 2026  |  Generated: 9 June 2026 05:00 GMT

To view the full AI prompts used to generate this digest check this file: 24h_2026-06-08_prompts.txt

Looking for older digests? Browse all digests here.