Daily Digest

AI (Artificial Intelligence)

During the past two weeks, the AI module made significant progress across core infrastructure, user experience, and ecosystem maturity, positioning the project closer to a stable 1.5.0 and 2.0.0 release.

Core Infrastructure and Architecture

The module laid important foundations for version 2.0. A long-running architectural refactor to adopt the Symfony AI Component (v0.8) as the platform provider advanced through multiple merge requests, introducing new inference provider, model router, and platform plugin types. Separately, the Automators subsystem was migrated to structured output, improving reliability when AI models generate field values. Token usage reporting was strengthened: the Fiber execution path—used for streaming chat—now correctly preserves token counts, and observability spans emit OpenTelemetry GenAI semantic conventions alongside existing ad-hoc metrics, so external monitoring tools can consume a standard shape. A new reranking processor for Search API was contributed, making the existing rerank operation type available to any search backend. Field Widget Actions integration with Automators was generalized and ported to the 2.x branch, reducing code duplication across widget types.

Quality, Testing, and Documentation

Multiple functional JavaScript tests were added for the AI CKEditor plugins, covering tone adjustment, translation, and summarization workflows. A commit-message validation rule was merged to enforce issue references on main branches without restricting feature-branch workflow. Several merge requests removed deprecated bundled submodules (AI Content Suggestions, Field Widget Actions) and clarified upgrade paths to their standalone contrib counterparts. Documentation received a last-modified timestamp via the mkdocs git-revision-date plugin, and planning began on a deterministic script to auto-generate the provider/operation-type matrix from live GitLab data.

User Experience and Configuration

The Prompt Library menu link was moved out from under Vector Database Configuration into the top-level AI Infrastructure section. The AI CKEditor status indicator ("AI Writer: Idle") became optional through a new settings toggle. Work is underway to allow chatbot access control to rely solely on Assistant entity roles, removing the redundant "access deepchat api" permission. A regression in the 1.4.x line—commits present in 1.3.3 but missing from 1.4.2—was identified and closed. The guardrail "RegExp" plugin label was renamed to "RegEx" for consistency. Several UX issues were reported and remain open: the chatbot closing on modal events, missing submit buttons in certain tool configuration forms, and rendering problems in dark Gin themes.

Risks and Blockers

No critical blockers were reported. A handful of open bugs—CKEditor context unavailable in off-canvas dialogs, taxonomy automator stripping HTML-encoded entities, non-streaming requests broken inside Fibers—have proposed fixes under review. The 1.5.x and 2.0.x branches remain in active development; no dates for stable releases were announced.

How can I help on this project?

• Champion the completion of the Symfony AI Component adoption (MR !1619 and its dependencies) by ensuring architectural review capacity and alignment on the 2.0 roadmap.
• Support ecosystem documentation work: the auto-generated provider matrix and the curated related-modules list both need resourcing to stay current as new providers and companion modules arrive.

Contributors: Abhisek Mazumdar (abhisekmazumdar) [1], Ahmad Khader (Ahmad-Khader), Ahmad Khalil (ahmad-khalil-imagex), AKHIL BABU (AkhilBabu) [1], Alphons Jaimon (AJV009), Ann Mary Sruthy (annmarysruthy), Arnaud Willame (arwillame), Artem Dmitriiev (a.dmitriiev) [3], Avinash jha (avinash.jha), Bryan Sharpe (b_sharpe), David Galeano (gxleano), Dimitar Rupov (mrupal), Eric Homanchuk (erichomanchuk), George Kastanis (zorz) [2], Hrishikesh Dalal (hrishikesh-dalal), Javier Destefanis (javierdestefanis), Joshua Fernandes (joshua1234511), Juan Martinez (jcmartinez), Jérôme Tchania (codeitwisely) [1], Levente Besenyei (LeventeB), m4olivei, Marcus Johansson (Marcus_Johansson) [2], Mukeysh, Omar Mohamad - El Hassan Lopesino (omarlopesino), Paul Sheldrake (paulsheldrake), Prabhavathi Vanipenta (prabha1997), Pravesh Poonia (Pravesh_Poonia), project update bot (Project-Update-Bot), Ricardo Castañeda (cadence96), Rob Loach (robloach), Ronald te Brake (ronaldtebrake), Scott Euser (scotteuser), scott falconer (scottfalconer), Shivam Sen (shivamsen_12579), Sujal Kshatri (sujal_31), Sven Decabooter (svendecabooter), Thomas Werthmüller (weri), velmir_taky (root_emarketing), Wolfgang Ziegler (fago)

View project on GitLab: ai

Drupal AI Initiative

The Drupal AI Initiative made significant progress across multiple workstreams during this period, with a strong focus on user experience design, demonstration infrastructure, and partnership growth.

A complete set of AI roadmap feature mockups was delivered, covering eight key editorial workflows including AI-assisted page creation, content review notifications, search, field automation, translation, and dashboard configuration. These interactive prototypes establish a shared design reference for the initiative's direction and were closed as completed during the period.

The foundation for an AI CMS Demo System was established, including repository setup, technical architecture, and proof-of-concept deployment. The demo will showcase AI capabilities built on Drupal CMS in a production-ready format, with hosting-specific configurations to support different deployment environments. This work closed as complete and provides a critical asset for showcasing the initiative's technical capabilities.

Planning advanced for an AI Content Review Workflow, with user experience design and technical architecture both in progress. Once delivered, this will enable editors to trigger AI-powered content quality checks, compliance validation, and brand consistency reviews directly within Drupal's editorial interface.

A new Gold partner, Reading Room, was onboarded during the period, expanding the initiative's contributor and funding base. The Enterprise AI Summit continued planning, with case study sourcing, sponsor updates, and mobile site improvements underway.

New concepts were introduced for a Cost & Usage module to give site owners visibility into AI spending and usage patterns, and for hosting provider skills that would allow developers to control deployment through conversational instructions.

How can I help on this project?

• Allocate resources to support the AI Content Review Workflow technical build now that design and architecture are defined, ensuring timely delivery of this flagship editorial feature.
• Champion adoption of the new demo system with key stakeholders and potential users to maximize visibility and feedback once deployment is complete.
• Approve partnership outreach to additional hosting providers beyond DevPanel to expand the reach of deployment automation skills.

View project on GitLab: ai_initiative

AI Integration - ECA

The AI Integration - ECA module is working toward a stable 1.0.0 release by expanding coverage of AI operations that can be automated through low-code workflows. The module currently supports five AI operation types; over the past two weeks, work has been organised to add actions for nine additional operations, including summarisation, text classification, text-to-image generation, audio-to-audio processing, image classification, image-to-image and image-to-video transformations, object detection, and speech-to-speech conversion. Proposals were also opened for simplifying how structured data formats are configured in chat workflows and for adding an interceptor submodule that will let administrators monitor or modify AI requests and responses as they flow through the system.

Three merge requests have been submitted covering text classification, summarisation, and a broader set of new operations. Automated compatibility updates for Drupal 12 are also in progress. All open issues are part of normal feature development for a pre-release module; no blocking concerns have been identified.

How can I help on this project?

• Ensure capacity is available for code review so the pending merge requests covering nine new operation types can be evaluated and merged promptly.
• Clarify the release timeline and any outstanding decisions needed to reach the 1.0.0 milestone.

Contributors: Joshua Fernandes (joshua1234511), Marcus Johansson (Marcus_Johansson), project update bot (Project-Update-Bot)

View project on GitLab: ai_integration_eca

AI Agents Test

AI Agents Test is a module that helps site builders validate whether their AI configurations are working as expected. Over the past two weeks, activity focused on planning improvements to testing workflows and clarifying the module's future direction.

A research issue was closed that explored whether this module should be merged with a related evaluation tool, ai_eval. This decision provides clarity on the project's scope and helps avoid duplication of effort across the Drupal AI ecosystem.

Work also progressed on a proposal to allow users to run more than one test group in bulk. This would make it easier for teams to test multiple AI models or configurations at once, rather than running tests individually. The capability would be particularly valuable when comparing different model performance or validating prompt engineering changes across several scenarios.

How can I help on this project?

• Confirm whether bulk test execution is a priority feature that warrants dedicated development resources, as it would improve efficiency for teams comparing multiple AI configurations.
• Clarify strategic positioning now that the module's scope has been separated from other evaluation tools in the ecosystem.

View project on GitLab: ai_agents_test

OpenAI Provider

The OpenAI Provider module made steady progress during this period, with two enhancements merged and three additional improvements under review.

Support for image-to-image capabilities was added, expanding the range of AI-powered image operations available to users. The default image generation model was also updated to remove the deprecated Dall-E 3 option, ensuring the module aligns with current provider offerings.

Three open merge requests are awaiting review: moderation skip functionality with test coverage, batch embeddings support for more efficient processing of multiple text inputs, and routine automated maintenance updates.

Overall, the module is progressing with incremental feature additions and maintenance work typical of active development.

How can I help on this project?

• Ensure reviewer capacity is available to assess and merge the three pending improvements, particularly the batch embeddings feature which may benefit users processing larger volumes of content.
• Consider whether formal testing resources could accelerate quality assurance for new features like moderation controls.

View project on GitLab: ai_provider_openai

Anthropic Provider

The Anthropic Provider module made significant progress during this period with the delivery of prompt caching capabilities. A merge request adding prompt caching and SDK improvements was completed and merged, giving administrators new tools to optimize performance and reduce costs. Users can now enable prompt caching through a simple toggle in the provider settings and select cache durations of either five minutes or one hour. This feature allows frequently-used prompts to be reused without reprocessing, potentially reducing API costs and improving response times for repetitive workflows.

Additionally, work is underway to address a compatibility issue with multi-agent conversations. A proposed fix for conversations ending with non-user messages remains open. This issue affects scenarios where AI agent handoffs result in message sequences that the Anthropic API does not accept, causing errors in tools like Drupal Canvas AI.

How can I help on this project?

• Prioritize review and testing of the open multi-agent conversation fix to unblock AI agent workflows and improve user experience.
• Allocate resources for documentation or user guidance on prompt caching benefits to drive adoption of the new cost-saving feature.

Contributors: Carlos Ospina (camoa), Rajab Natshah (RajabNatshah)

View project on GitLab: ai_provider_anthropic

Fireworks AI

The Fireworks AI module saw focused maintenance activity during this period. Three commits were made to address requirements and improve the module's underlying structure. The work included fixes to module requirements and a refactoring of the rerank functionality to use standardized base components, which should improve consistency and maintainability going forward.

No issues were updated and no merge requests were completed during this period, indicating a quiet fortnight focused on incremental improvements rather than major feature development or bug resolution.

How can I help on this project?

• Clarify the project's strategic priority and resource allocation to determine whether this module should receive sustained development effort or remain in maintenance mode.
• Support community engagement by encouraging user testing and feedback to identify the most valuable next features or improvements.

Contributors: Marcus Johansson [3]

View project on GitLab: fireworksai

MCP Client

The MCP Client module saw minimal development activity during the reporting period. An automated maintenance update was initiated to keep the project aligned with current standards, though no work was merged or completed during this time.

The module remains in an early stage of development, and the lack of substantial progress may indicate limited active development resources or that work is occurring outside the main repository.

How can I help on this project?

• Clarify whether this module remains a strategic priority and allocate dedicated development resources if continued progress is expected.
• Identify and engage potential contributors or partners who could accelerate development and bring the module to a stable release.

Contributors: project update bot (Project-Update-Bot)

View project on GitLab: mcp_client

amazee.ai Private AI Provider

This module provides integration with amazee.ai's private AI infrastructure. During the reporting period, two fixes were merged to address compatibility and configuration issues.

A missing database extension dependency was resolved, ensuring the module correctly declares its requirement for PostgreSQL functionality. Separately, validation was improved to prevent configuration errors when key identifiers are empty or invalid.

An automated compatibility update for Drupal 12 has been prepared and is awaiting review. This work ensures the module will remain functional as the platform evolves.

The module continues to mature with focused bug fixes that improve reliability and deployment safety. No blocking issues were reported during this period.

How can I help on this project?

• Review and approve the pending Drupal 12 compatibility update to ensure timely platform support.
• Confirm ongoing resourcing for maintenance as the module moves toward stable release.

View project on GitLab: ai_provider_amazeeio

Gemini Provider

The Gemini Provider module saw focused development activity over the past two weeks, advancing key capabilities for the platform's AI integration.

Work progressed on expanding the module's embedding functionality. A proposal for batch embeddings support is under review, which would allow the system to process multiple embedding requests more efficiently. Separately, development continues on image embeddings, which would enable the platform to index and search using images—a differentiating capability that few other providers offer.

Two issues address module stability. A fix for content blocking error handling is in progress, which would prevent the system from crashing when content safety filters are triggered. The team is also working through code quality checks to ensure the automated test suite accurately reflects the module's health. Additionally, automated compatibility updates for Drupal 12 are being prepared.

The module remains in active development with normal issue activity for this stage.

How can I help on this project?

• Prioritize which embedding capability—batch processing or image support—delivers greater business value, to help the team sequence work.
• Ensure testing resources are available to validate content blocking scenarios before release.

Contributors: Artem Dmitriiev (a.dmitriiev), Ishani Patel (IshaniPatel), Joshua Fernandes (joshua1234511), project update bot (Project-Update-Bot)

View project on GitLab: gemini_provider

Huggingface Provider

The Huggingface Provider module saw focused development activity during this period. A merge request was opened to add extractive question answering capability to the module. This feature would enable the platform to automatically answer questions by extracting relevant information from provided text, expanding the module's natural language processing capabilities beyond its current functionality.

The module remains in active development with no critical blocking issues reported during this period. As an integration module connecting Drupal to Huggingface's machine learning services, steady feature additions like this help broaden the range of AI-powered content and user experience features available to site builders.

How can I help on this project?

• Provide clarity on which AI capabilities are strategic priorities for upcoming releases to help focus development effort.
• Ensure the team has access to stakeholders who can validate whether new features like question answering meet real user needs.

View project on GitLab: ai_provider_huggingface

Microsoft Azure AI

The Microsoft Azure AI module saw focused development activity during this period, with improvements made to the text-to-image capabilities. This enhancement enables users to generate images from text descriptions more effectively, supporting content creation workflows and creative automation use cases.

The module continues to evolve as an integration point between Drupal sites and Microsoft's Azure AI services, allowing organizations to leverage cloud-based artificial intelligence features for content generation and media production.

How can I help on this project?

• Consider identifying internal use cases that would benefit from AI-generated imagery to validate the module's value proposition.
• Evaluate whether additional resourcing could accelerate development toward a stable release suitable for production deployment.

Contributors: Marcus Johansson [1]

View project on GitLab: ai_provider_azure

Milvus VDB Provider

The Milvus VDB Provider project saw preparatory work this period, with no changes merged into the codebase. Two merge requests remain open for review.

A batching feature for indexing operations has been proposed. This enhancement would allow the system to process multiple indexing operations together rather than one at a time, improving efficiency when working with large volumes of data. This capability is particularly valuable for organizations managing extensive content libraries or knowledge bases.

Additionally, an automated maintenance update has been prepared to keep the project aligned with Drupal coding standards and infrastructure requirements.

Both changes are awaiting review before they can be incorporated. With no commits merged during this period, there are no immediate changes to functionality or stability.

How can I help on this project?

• Ensure reviewer capacity is available to evaluate the pending batching enhancement, which could unlock performance improvements for users with large-scale indexing needs.
• Clarify strategic priorities for the module's feature roadmap to help contributors focus effort on high-value capabilities.

Contributors: Artem Dmitriiev (a.dmitriiev), project update bot (Project-Update-Bot)

View project on GitLab: ai_vdb_provider_milvus

Postgres VDB Provider

During this period, a merge request was opened to address dependency management within the module. The work aims to remove the hard requirement on a specific database extension, which would provide more flexibility in how the module can be deployed and integrated into different environments. This change could reduce implementation barriers and make the module more adaptable to varying infrastructure setups.

No merge requests were completed during this period, and no open issues required attention. As an early-stage module, this level of activity is typical, with development progressing incrementally.

How can I help on this project?

• Ensure the team has capacity to review and finalize the pending dependency work, which could unlock broader adoption scenarios.
• Clarify strategic priorities for this module relative to other vector database integrations to guide future resourcing decisions.

Contributors: Emiel Bloem (emielb)

View project on GitLab: ai_vdb_provider_postgres

AI Image Alt Text

The module saw limited activity during this period, with one merge request opened to expand compatibility with newer versions of the AI module framework. This work aims to ensure the module can integrate with the latest AI infrastructure, which will help maintain its viability as the broader AI ecosystem evolves.

No issues were updated, no code was committed, and no merge requests were finalized during this two-week window. This suggests the project is in a quiet phase with ongoing but not yet completed work.

How can I help on this project?

• Assign capacity to review and finalize the AI module compatibility update to keep the module current with its dependencies.
• Consider whether this module aligns with strategic priorities and warrants continued investment or formal release planning.

Contributors: Peter Philipp (daspeter)

View project on GitLab: ai_image_alt_text

Model Context Protocol

The Model Context Protocol module is receiving automated compatibility updates to prepare for Drupal 12. A compatibility issue was opened to track the removal of deprecated code, with a corresponding merge request containing proposed fixes. No changes have been merged yet during this period.

This automated maintenance work ensures the module will continue functioning when Drupal 12 is released, reducing technical debt and avoiding future migration costs. The updates are being generated systematically to identify code that will no longer be supported in the next major version of Drupal.

How can I help on this project?

• Allocate developer time to review and test the pending compatibility updates, ensuring the module is ready before Drupal 12's release timeline becomes critical.
• Clarify the module's strategic importance and maintenance priority to help the team decide whether to invest in Drupal 12 support or sunset the project.

Contributors: project update bot (Project-Update-Bot)

View project on GitLab: mcp

AI Agents

The AI Agents module made steady progress during this period, with work focused on improving usability, fixing structural issues, and preparing for broader platform integration.

Several administrative refinements were addressed. The module's menu structure is being aligned with the updated AI 1.4.x navigation, moving AI Agents under a new "Tools & Automation" category to improve discoverability. A schema was added for the CKEditor integration to eliminate configuration warnings. Work also began to improve the structured-output schema editor, replacing a plain text field with a purpose-built JSON editor that provides validation and syntax highlighting.

Multiple bug fixes addressed runtime reliability. Issues that previously broke the user interface when tools encountered errors are now handled more gracefully. A fix was completed to ensure that tools marked as "return directly" properly track their identifiers in the conversation history. Work is in progress to resolve how Views agent updates are applied to non-default displays, ensuring that configuration changes appear correctly at runtime.

Strategic development included logging interface documentation and exploration of a plugin system to export or synchronize agents to external platforms such as Claude Code and VAPI, which would allow Drupal-configured agents to operate in external AI ecosystems. Work also continued on integrating agents with the chat processor architecture, enabling tighter parity with the AI Assistant layer.

Contributors: Ahmad Khader (Ahmad-Khader), Bharat Kelotra (bharatkelotra), David Galeano (gxleano), Jürgen Haas (jurgenhaas), Marcus Johansson [1], Nicolae Procopan (thebumikgmail.com), Prabhavathi Vanipenta (prabha1997), project update bot (Project-Update-Bot), Rajab Natshah (RajabNatshah), Sven Decabooter (svendecabooter)

How can I help on this project?

• Clarify priority and scope for the external platform integration work to inform resourcing decisions and avoid scope creep in a maturing codebase.
• Designate ownership for the Views agent display override issues to unblock validation and merge of fixes already in review.

View project on GitLab: ai_agents

Tool API

The Tool API module, which enables AI agents to invoke structured operations within Drupal, saw focused refinement work over the past two weeks. A critical usability issue was resolved: the command-line interface had been masking failure messages when tools encountered errors, making it difficult to understand what went wrong. This bug was fixed, ensuring meaningful error messages now surface correctly when operations fail.

Discussion continued on several enhancements that would improve how the module integrates with AI systems and developer workflows. Teams explored File Input Handling to understand how tools can securely accept and process files from AI agents. Work also progressed on Add skills for generating Tools, which was closed during the period, aiming to make it easier for AI coding assistants to generate new tool definitions. Proposals were discussed around Allow controlling Tool return format to support both human-readable and machine-readable outputs, and Proposal: make Tool's declarations introspectable over the CLI to help consumers better understand available tools before invoking them.

A serialization compatibility issue affecting REST endpoints received attention, with a merge request now under review. Additional planning issues remain open around preview capabilities and improved tool discovery for large-scale AI agent deployments.

How can I help on this project?

Consider allocating resources to review and finalize the serialization fix, which affects REST API integration. Support decision-making around tool discovery and execution patterns to clarify the roadmap for AI agent scaling. Advocate for user testing of the file input handling research once prototype work completes.

View project on GitLab: tool

Drupal Canvas

During the reporting period, substantial progress was made on symmetric translation support for multilingual Canvas sites. A large merge introduced automatic propagation of component version updates to all translations of both content and config entities, preventing translation drift when code components evolve. The underlying data model now treats content and config translations uniformly through a new staged-config-override abstraction. Several bugs blocking the translation milestone were resolved, including a fatal error when an image reference pointed to a deleted file, and incorrect translatability of entity-reference inputs. The pending-changes review UI was refined to prevent accidental per-translation operations until the full langcode-aware publishing flow is in place.

Work continued on conflict detection and resolution. The back-end now recognizes when an auto-save diverges from a published entity that has changed externally, and the UI displays conflict indicators during selective publishing. This functionality is currently behind a feature flag while the resolution flow is refined. A rudimentary resolution interface was merged, offering side-by-side comparison and viewport selection.

Canvas AI improvements included a fix to the orchestrator agent's UUID persistence during post-updates, correct exposure of block component props to agents, and test fixtures for page-builder and template-builder agents. Several prompt bugs affecting component generation (deprecated imports, prop/slot collisions, hardcoded errors) were consolidated and resolved. In-progress chat messages now persist when page metadata changes mid-run.

The CLI was updated: deprecated upload and download commands were removed, media reconciliation now respects sync settings, and command output was cleaned up for improved readability and AI-agent consumption. JavaScript packages were updated, and the kernel test suite was split across four parallel CI jobs to reduce pipeline time. A regression causing boolean props to auto-enable when other fields changed was traced to falsy-value coercion and a Twig template quirk, and fixed. A full-screen "page" extension type was introduced, and test coverage for multi-translation component-instance updating was strengthened.

How can I help on this project?

• Prioritize review and feedback on the conflict-resolution feature flag and UI (still in draft/refinement) so it can move to stable and support production editorial workflows.
• Allocate resources to expand test coverage for content-entity-reference props and translation edge cases, reducing manual QA overhead as the feature set grows.

Contributors: **** (project_19391_bot_cb443e60680723bba08a5351bf595251), Abhishek Dhariwal (adaptative) [1], Adam G-H (phenaproxima), Ahmad Khalil (ahmad-khalil-imagex) [1], AKHIL BABU (AkhilBabu) [3], Alamgir Hossain (alamgir.hossain.cse), Andrew Belcher (andrewbelcher), Ben Mullins (bnjmnm) [3], Björn Brala (bbrala) [1], Bálint Kléri (balintbrews) [7], Chandan Singh (chandu7929), Christian López Espínola (penyaskito) [8], Feliksas Mazeikis (f.mazeikis) [2], Gábor Hojtsy (goba), Jakob P (japerry), Kunal Sachdev (kunal.sachdev) [2], Lauri Timmanee (lauriii) [1], Matt Glaman (mglaman) [2], Narendra Singh Rathore (NarendraR) [1], Oleksandr Antonchyk (Anton4yk), Priyanka Patel (priyankapatel), project update bot (Project-Update-Bot), Rajab Natshah (RajabNatshah), Ravi Maniyar (ravi.maniyar.123) [2], Sally Young (justafish), scott falconer (scottfalconer), Shubham Prakash (shubham.prakash), Ted Bowman (tedbow) [1], utkarsh_33 (Utkarsh_33) [1], Wim Leers (wimleers) [7], wotnak

View project on GitLab: canvas

Context Control Center (CCC)

Context Control Center shipped significant progress toward its beta3 milestone during this period, with a focus on documentation, UI improvements, and stabilizing core features ahead of the first release candidate.

Several important enhancements were merged to improve the user experience. Full view layout for context item pages was completed, replacing the default field list with a two-column design that surfaces scope and subcontext details more clearly. The context listing page was refined to merge separate Use Cases and Scope columns into a unified display with cascading filters, making it easier to browse and filter context items. The overview page description and subcontext feature documentation were updated to clarify their purpose and benefits.

A new experimental submodule, Document Loader: Context Importer, was introduced to allow users to import context from PDF, Markdown, text files, and URLs via an MDXEditor integration. Example context items from the Chicago Driesnote demo were added to help testers explore the module's capabilities.

Important API and configuration work included validation to prevent invalid global context limits, better support for scope value extension via hooks, a public API for scope subscription forms, and making the Scheduler feature optional. Conditional subcontext received a toggle, caching, and per-request limit to manage latency and provider costs. Usage tracking query inefficiencies were resolved by batching database operations.

Quality improvements included PHPUnit process isolation attributes on all kernel tests, eslint error fixes, and bundleless context item follow-ups. Developer resources were strengthened with GitLab issue and merge request templates, a SECURITY.md file, and API stability documentation distinguishing supported from internal interfaces.

Beta testing recruitment is underway, and planning for the beta3 release continues with remaining milestone work focused on scope plugin refinement, documentation consistency, and accessibility improvements identified through AI Best Practices review.

How can I help on this project?

• Support recruitment and onboarding of beta testers by sharing CCC resources and demo videos with relevant stakeholder groups or partner organizations.
• Allocate design and accessibility review time to help resolve open UX issues flagged for the release candidate.
• Help prioritize scope plugin and API stabilization work against other organizational commitments to keep the beta3 and RC1 timeline on track.

Contributors: Abhisek Mazumdar (abhisekmazumdar) [2], Ahmad Khader (Ahmad-Khader) [1], Anikó Viola (violaniko) [1], Daniel Rodriguez (danrod) [1], Eric Homanchuk (erichomanchuk) [1], Jesse Hofmann-Smith (jessehs), Joshua Fernandes (joshua1234511), Juan Correa (jucs7) [2], Kristen Pol (kepol) [9], Matt Glaman (mglaman) [5], Pravesh Poonia (Pravesh_Poonia), project update bot (Project-Update-Bot), Rob Loach (robloach) [2], Tamas Balog (tbalog)

View project on GitLab: ai_context

AI Dashboard

The AI Dashboard project made progress on documentation and user experience improvements during this period. Work continued on expanding the AI ecosystem directory, which will help site administrators discover and browse available AI-related modules. A modal display issue affecting the recipe installation feature was identified and postponed for future resolution.

The most significant development is an open proposal to Improve AI Ecosystem, which includes plans to integrate a module directory directly into the Dashboard interface, allowing users to browse AI modules without leaving their site. The work also includes improving category detection and adding automated updates to keep the ecosystem listing current. This would streamline the module discovery process and reduce friction for teams looking to expand their AI capabilities.

A bug affecting the recipe installation modal was reported and postponed, which currently prevents users from installing AI features through the interface as intended.

How can I help on this project?

• Clarify whether the recipe installation feature is a priority for users, or if postponing this capability is acceptable for current adoption goals.
• Review the ecosystem integration proposal to ensure it aligns with your broader module distribution and user onboarding strategy.

Contributors: Rob Loach (robloach)

View project on GitLab: ai_dashboard

AI Search

The AI Search module made progress on indexing performance and search filtering capabilities during this period. A feature enabling long-running chunk indexing was completed and merged, allowing the system to handle larger content sets without breaking changes to existing installations. Work is underway on batch embeddings to further improve indexing speed.

A new approach to dynamic search filtering is in development through the FilterableRagTool feature, which would allow sites to filter search results by custom attributes (such as case studies or content types) at query time. An existing filter issue was closed during the period.

Four configuration and error-handling issues were opened late in the period, highlighting gaps in logging, parameter handling, and support for non-embedding search providers. These are typical of alpha-stage development and will need attention to improve reliability and clarity for site administrators configuring the module.

How can I help on this project?

• Prioritize resources to resolve the four recently opened configuration and error-handling issues, which affect admin experience and provider flexibility.
• Clarify the roadmap for the 2.0 stable release, particularly around compatibility with the AI Core 2.x integration work that is currently blocked.

View project on GitLab: ai_search

Field Widget Actions

Field Widget Actions enables users to add action buttons (such as AI content generators) directly into content editing forms. Over the past two weeks, the project delivered several important refinements that improve usability, compatibility, and the reliability of AI-powered workflows.

Three fixes were completed and merged. A bug preventing unavailable actions from appearing in configuration menus was resolved, reducing confusion during setup. A compatibility issue affecting integrations like autosave was addressed so that when users select AI-generated suggestions, other form elements respond correctly. Automated compatibility updates for the upcoming Drupal 12 release were also applied.

A significant new capability was delivered: interactive refinement for AI content. Users can now iteratively improve generated content through a modal interface, submitting follow-up prompts to refine results before inserting them into the field. This feature is optional and does not affect existing workflows.

Three enhancements remain in progress. Work is underway to simplify how automated configuration recipes declare actions, to support additional third-party field widgets, and to extend action buttons to checkbox and radio inputs. A minor issue regarding button label translation was reported and is under review.

How can I help on this project?

• Ensure the configuration recipe enhancement receives review and testing resources so automated deployments can be simplified.
• Support testing and sign-off for the checkbox and radio button enhancements currently in review.
• Prioritize resolution of the translation workflow gap if multilingual deployments are planned.

Contributors: Artem Dmitriiev (a.dmitriiev) [1], Daniel Bosen (daniel.bosen), Edward Zwart (ezeedub) [1], project update bot (Project-Update-Bot) [1], Scott Euser (scotteuser), Volker Killesreiter (volkerk) [1]

View project on GitLab: field_widget_actions

AI translate

The AI translate module saw focused attention on content quality and scale challenges during this period. Work is underway to address timeouts when translating large pages (over 1,000 words) by splitting content into manageable chunks. A merge request has been prepared to resolve 504 Time out on large content, which should enable translation of larger documents without service interruptions.

Two significant content integrity issues were flagged: HTML markup is corrupted/not preserved when translating formatted text fields and HTML markup gets added after certain tags. These affect the reliability of translations for content containing rich text formatting, which is common in most content management scenarios.

Strategic development discussions include a proposal to batch field translations into single requests rather than translating each field separately, which would reduce redundant API calls and associated costs. A workflow refactoring plan was raised to support agent-driven translation. Compatibility work for Drupal 12 progressed with an automated update.

How can I help on this project?

• Prioritize resolution of the HTML markup preservation issues, as they affect content quality and may require prompt engineering or architectural decisions about how formatted content is processed.
• Review and approve the batching proposal for field translation to reduce API costs and improve efficiency before implementation begins.

Contributors: Miriam Nugent (Peacog), project update bot (Project-Update-Bot)

View project on GitLab: ai_translate

Pinecone VDB Provider

The Pinecone VDB Provider module, which enables integration with Pinecone vector database services, saw focused development activity during this period. Work centered on improving the efficiency of data indexing operations through a new batching capability, inspired by similar improvements in a related module.

Two proposals were prepared to add batching support for indexing operations, one targeting the current stable version and another for the next major release. This enhancement will allow the module to process multiple data items together rather than individually, which should reduce processing time and improve performance when handling large volumes of content. Both proposals are awaiting review.

No code was merged or deployed during this period. The module remains in active development with work progressing on performance optimization features.

How can I help on this project?

• Allocate technical review capacity to evaluate the pending batching proposals, enabling the performance improvements to move forward.
• Clarify versioning strategy: determine whether batching should be introduced in the stable release, the next major version, or both.

Contributors: Rob Loach (robloach)

View project on GitLab: ai_vdb_provider_pinecone

AI powered Search Block

Development activity over the past two weeks focused on feature enhancement and maintenance. Work began on adding guardrail set selection capabilities to the search block, which will give administrators better control over content safety and quality standards in AI-generated search results. This addresses a need for more configurable content moderation in user-facing search experiences.

Two automated maintenance updates were also opened to keep the module's dependencies and coding standards current. No code was merged during this period, indicating work remains in review and validation stages.

How can I help on this project?

• Allocate review resources to assess and approve the pending guardrail selection feature and maintenance updates, accelerating their path to release.
• Clarify strategic priorities for content moderation features to guide decision-making on the guardrail implementation approach.

Contributors: Abhisek Mazumdar (abhisekmazumdar), project update bot (Project-Update-Bot)

View project on GitLab: ai_search_block

FlowDrop

FlowDrop has begun preparing for compatibility with the upcoming Drupal 12 release. An automated compatibility fixes issue was opened to track removal of deprecated code that will no longer be supported in the next major version of Drupal. A corresponding merge request has been prepared containing the necessary updates.

This work ensures FlowDrop will continue to function when Drupal 12 is released, preventing disruption for sites that upgrade their core platform. The changes address technical deprecations introduced in earlier Drupal versions that must be resolved before the next major release.

No commits were finalized during this period, indicating the compatibility changes are still under review. This is normal preparation work for upcoming platform changes and does not represent any immediate risk to current users.

How can I help on this project?

• Allocate developer time to review and finalize the Drupal 12 compatibility updates so the module is ready before the core platform release.
• Clarify the project's timeline and priority for Drupal 12 support to guide resource planning.

Contributors: project update bot (Project-Update-Bot)

View project on GitLab: flowdrop

FlowDrop UI Agents

Work continued during this period to prepare the module for compatibility with Drupal 12. An automated compatibility update was opened to address deprecated code patterns that will no longer be supported in the next major version of Drupal. A corresponding merge request was created with proposed changes spanning over 230 lines of code.

This compatibility work is essential groundwork to ensure the module remains functional when Drupal 12 is released. Without these updates, the module would eventually break for users upgrading their systems. The changes are still under review and have not yet been integrated.

No commits were merged during this period, meaning no new functionality was delivered to users. The focus remains on ensuring long-term platform compatibility rather than adding features or fixing user-facing issues.

How can I help on this project?

• Allocate technical review capacity to evaluate and merge the pending Drupal 12 compatibility updates, preventing delays in platform readiness.
• Clarify the module's strategic priority and roadmap to determine whether active development should continue or resources should be redirected.

Contributors: project update bot (Project-Update-Bot)

View project on GitLab: flowdrop_ui_agents

Langfuse

Activity during this period was limited to automated maintenance. A merge request containing automated fixes was opened by the project update bot. This type of work helps keep the module aligned with Drupal coding standards and platform updates, reducing technical maintenance burden over time. No issues were updated and no code was committed to the main codebase during this window.

The module appears to be in a quiet phase with minimal development activity. With only automated maintenance in progress and no active issue resolution or feature development, the project may benefit from renewed attention to advance its roadmap.

How can I help on this project?

• Review and approve the pending automated maintenance work to keep the module's infrastructure current.
• Assess whether dedicated developer time should be allocated to move the project toward a stable release.
• Clarify the strategic priority of this module and communicate direction to potential contributors.

Contributors: project update bot (Project-Update-Bot)

View project on GitLab: langfuse

Orchestration

The Orchestration module saw planning and design activity across three strategic initiatives during this period, though no code was merged. Work is underway to expand integration options and improve the module's flexibility.

Progress centered on integration with n8n, a workflow automation platform. This will enable two-way communication between Drupal and n8n, allowing Drupal events to trigger external workflows and n8n to create or update Drupal content in return. For organizations already using n8n, this would reduce manual bridging work and open new automation possibilities.

A proposal to adopt JSON Schema for defining inputs, outputs, and configuration is being explored. This would provide clearer contracts for how workflow components connect, making it easier to validate and document automations before they run.

Separately, work is planned to allow ECA workflows to invoke Orchestration services. Currently, Orchestration can call ECA, but not the reverse. Closing this gap would let teams use ECA's event system to trigger any registered Orchestration service.

How can I help on this project?

• Clarify strategic priority between n8n integration and the JSON Schema refactor to focus development effort.
• Confirm whether n8n is the right platform investment based on organizational automation roadmap and existing tooling.

View project on GitLab: orchestration

AG-UI

Over the past two weeks, work began on preparing AG-UI for compatibility with the upcoming Drupal 12 release. An automated compatibility fixes issue was opened to track the removal of deprecated code patterns, with an associated merge request now under review. This proactive maintenance ensures the module will continue functioning when Drupal 12 becomes available, reducing future technical debt.

Separately, planning continued on the Drupal Assistant Implementation meta issue, which aims to evolve the current proof-of-concept assistant into a production-ready feature that could replace the existing chatbot functionality. This represents a strategic enhancement to how users interact with AI capabilities within Drupal.

No code was merged during this period, reflecting typical early-stage development activity where compatibility groundwork and feature planning are underway.

How can I help on this project?

• Prioritize whether the Drupal Assistant feature should be resourced for production development or remain experimental, enabling clearer roadmap planning.
• Ensure the team has capacity to review and merge the Drupal 12 compatibility changes before the next major Drupal release.

Contributors: project update bot (Project-Update-Bot)

View project on GitLab: agui

AI Related Content

During this period, a significant defect affecting the Pinecone integration was resolved. The issue prevented the related content view from returning any results when using Pinecone as the vector database provider. The underlying problem involved a query filter that was being silently ignored, causing the system to retrieve random semantic matches rather than properly filtered content. This has now been fixed, restoring the module's ability to deliver relevant content recommendations to end users.

Additionally, automated Drupal 12 compatibility work has begun, with an open merge request containing compatibility updates. This positions the module to support the next major version of Drupal when it becomes available.

How can I help on this project?

• Prioritize review and approval of the pending Drupal 12 compatibility updates to ensure the module remains supportable as the platform evolves.
• Consider allocating testing resources to validate the Pinecone fix across production-like environments before wider deployment.

Contributors: project update bot (Project-Update-Bot), Scott Euser (scotteuser) [1]

View project on GitLab: ai_related_content