> ## Documentation Index
> Fetch the complete documentation index at: https://docs.flowx.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# FlowX.AI 5.10.0 Release Notes

> What's new in FlowX.AI 5.10: PostgreSQL and SOAP data sources, AI Developer script-node code generation, file uploads in chat-driven apps, and AI and document-intelligence improvements.

{/* HERO: scripts/hero-5.10.0.html — glimpses built for the locked triad (PostgreSQL data source, AI Developer script-node code, chat file upload); verified 3-up at the 718px Mintlify column + token audit clean. DEPLOYED (Dragos) to docx.flowx.ai/5.9/hero-5.10.0.html — live (200). Font src is same-level OpenSans-*.ttf → /5.9/; prefixed TTFs 403 to anon so type falls back to system Open Sans cross-origin (known/accepted, per feedback_hero_production_fidelity). Fixed-height embed per hero skill step 9 (Mintlify strips <style> + force-injects 16/9; inline style is the only reliable lever). */}

<div style={{position: "relative", width: "100%", height: "720px", borderRadius: "12px", overflow: "hidden"}}>
  <iframe src="https://s3.eu-west-1.amazonaws.com/docx.flowx.ai/5.9/hero-5.10.0.html" title="FlowX.AI 5.10 release hero" style={{position: "absolute", top: 0, left: 0, width: "100%", height: "100%", border: "none"}} />
</div>

<Info>
  **FlowX.AI 5.10 is a SaaS release.** New capabilities are available on managed (SaaS) deployments; self-hosted deployments receive them with the next LTS release.
</Info>

<Note>
  **Deployment models:**

  * <Badge color="blue" icon="cloud">SaaS</Badge> Managed FlowX cloud deployments
  * <Badge color="green" icon="globe">All Deployments</Badge> Available in both self-hosted and SaaS
</Note>

FlowX.AI 5.10 is a SaaS-managed release: FlowX handles deployment-level changes such as infrastructure, Kafka topics, and service configuration. This release adds PostgreSQL and SOAP data sources, multi-file upload in chat-driven apps, and a set of AI and document-intelligence capabilities, alongside broader audit coverage, roles that travel with copied resources, and workflow data-partitioning improvements.

***

**What's new?**

<Tabs>
  <Tab title="Core & Platform">
    🗂️ [**Expanded audit logging**](#expanded-audit-logging) - Broader coverage of access and lifecycle actions\
    📑 [**Copy with roles**](#copy-with-roles) - Roles travel with copied resources\
    ⚡ [**Platform performance and fixes**](#platform-performance-and-fixes) - Task-manager user fetch, project versioning, export/merge
  </Tab>

  <Tab title="Designer & Renderer">
    💬 [**Chat file upload**](#chat-file-upload) - Attach multiple files in conversational apps\
    ⬇️ [**File download component**](#file-download-component) - Download files from a rendered app (web and Android)\
    🔀 [**Conditional event-handler triggers**](#conditional-event-handler-triggers) - Fire event handlers based on a condition\
    🎨 [**Image color property**](#image-color-property) - Color control on the IMAGE component\
    💾 [**Save local variable**](#save-local-variable-action) - Capture a collection row into local data on click\
    🪟 [**Global header and footer**](#global-header-and-footer) - Consistent header and footer across a UI Flow\
    🧩 [**Reusable UI templates**](#reusable-ui-templates) - Action mapping, undo/redo, and dependency-version fixes
  </Tab>

  <Tab title="Integrations & Data">
    🐘 [**PostgreSQL data source**](#postgresql-data-source) - Connect workflows directly to PostgreSQL and run SQL queries\
    🧼 [**SOAP integration**](#soap-integration) - Import a WSDL and call SOAP operations from workflows\
    🧮 [**FlowX Database aggregation**](#flowx-database-aggregation) - MongoDB aggregation pipelines\
    🗄️ [**Documents tab in database testing**](#documents-tab-in-database-testing) - Inspect data while testing DB operations
  </Tab>

  <Tab title="AI & Documents">
    📊 [**Tabular data extraction node**](#tabular-data-extraction-node) - Read spreadsheets and CSV into workflow data\
    🔁 [**Knowledge Base error handling**](#knowledge-base-error-handling-and-retry) - Retry failed KB ingestion\
    ✨ [**AI Developer: script-node code**](#ai-developer-generate-script-node-code) - Generate node code from a prompt\
    🧪 [**AI evaluation experiments**](#ai-evaluation-experiments) - Multiple datasets and improvement suggestions\
    🛡️ [**AI node output reliability**](#ai-node-output-reliability) - Schema validation with auto-repair and clearer run errors\
    📝 [**Template Filling node**](#template-filling-node) - Generate documents from templates in workflows\
    🗂️ [**File manipulation nodes**](#file-manipulation-nodes) - Combine and split files in workflows\
    📄 [**DOCX document generation**](#docx-document-generation) - Generate documents as DOCX in addition to PDF\
    🏷️ [**Document Tag**](#document-tag-renamed-from-document-type) - "Document Type" renamed\
    🔑 [**Mistral model provider**](#mistral-model-provider) - Bring your own Mistral API key
  </Tab>
</Tabs>

### **Expanded audit logging**

<Card title="Broader audit coverage for access and lifecycle actions" icon="clipboard-list" iconType="duotone">
  Audit logging now records a wider set of actions: **sharing**, **user invitations**, **designer-user, end-user, and group management**, and actions on **process instances** and **UI flows**. Adding a dependency is now audited, and terminating process, workflow, or UI-flow instances is captured as well.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **Copy with roles**

<Card title="Roles travel with copied resources" icon="copy" iconType="duotone">
  Copying a process or UI flow to another project now also copies the resource's associated **roles**, so access configuration no longer has to be recreated by hand after a copy.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **Platform performance and fixes**

<Card title="Platform performance and fixes" icon="wrench" iconType="duotone">
  <Tabs>
    <Tab title="Performance">
      The task manager fetches users through the authorization-system instead of calling Keycloak directly, cutting per-request identity lookups. {/* CORE-4676 verified origin/main 5f8ea80d26. CORE-4675 = sibling, covered by 4676 (no separate commit found 06-30). */}

      MongoDB partition archiving scales for large datasets, with partition-ID backfill for legacy data and archiving fixes on PostgreSQL and Oracle. {/* PLAT-1728/1729 verified origin/release/5.9.x; PLAT-1727 verified origin/main 848c2c0245 */}
    </Tab>

    <Tab title="Bug fixes">
      <Accordion title="Core platform fixes">
        * **Project selector:** projects without a work-in-progress branch are correctly shown as unavailable. {/* CORE-3998 verified origin/main fc03190b07 */}
        * **Start Solution:** no longer fails with "no entry point set" until the page is refreshed. {/* CORE-4823 */}
        * **Export and merge:** fixed errors when exporting a project build with imported libraries and when merging project versions, including a 500 during merge. {/* CORE-4934, CORE-4939, CORE-4944 */}
        * **Merge-conflict checks:** no longer fail on processes that contain conflicts. {/* CORE-4924 */}
        * **Allocation roles:** candidate-user search works for users without runtime permissions on the asset. {/* CORE-4929 verified origin/main 895d62d886 */}
        * **Version switching:** the selected resource opens when changing app versions, and reusable-function code refreshes on version switch. {/* CORE-4901, CORE-4911 */}
        * **Process engine:** resolved an incident-save deadlock. {/* PLAT-1797 */}
        * **FlowX Database:** inserting a document with a reserved `_id` key is now rejected with a clear error instead of breaking later reads. {/* MT-3025 */}
      </Accordion>
    </Tab>
  </Tabs>

  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **PostgreSQL data source**

<Card title="Connect to PostgreSQL databases from workflows" icon="database" iconType="duotone">
  <Tabs>
    <Tab title="Overview">
      Integration Designer adds **PostgreSQL** as a data source type, alongside Oracle. Connect a workflow directly to a PostgreSQL database and run SQL queries against it as a workflow node, without a custom REST wrapper. See the [PostgreSQL Database](/5.9/docs/platform-deep-dive/integrations/postgresql-database) documentation.
    </Tab>

    <Tab title="Key Features">
      <CardGroup cols={2}>
        <Card title="Direct connection" icon="plug">
          Configure host, port, database, schema, and credentials, with an optional SSL connection and a **Test connection** check before saving.
        </Card>

        <Card title="SQL query operations" icon="terminal">
          Define SQL query operations on the data source and call them from workflow nodes.
        </Card>

        <Card title="Connection pool tuning" icon="sliders">
          Advanced settings cover pool size, connection and query timeouts, keep-alive, and a validation query.
        </Card>

        <Card title="Familiar experience" icon="copy">
          Shares the same configuration and operations experience as the existing Oracle data source.
        </Card>
      </CardGroup>
    </Tab>
  </Tabs>

  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **SOAP integration**

<Card title="Call SOAP web services from workflows" icon="file-code" iconType="duotone">
  <Tabs>
    <Tab title="Overview">
      Integration Designer adds a **SOAP System** data source. Import the service's **WSDL** — from a URL or an uploaded file — and FlowX discovers the operations with their input, output, and fault schemas, composes the SOAP envelope at runtime, and lets you call each operation from workflows through a **SOAP Operation** node. See the [SOAP System](/5.9/docs/platform-deep-dive/integrations/soap-system) documentation.
    </Tab>

    <Tab title="Key Features">
      <CardGroup cols={2}>
        <Card title="WSDL import" icon="file-import">
          Preview and select the operations to import; re-import refreshes schemas while preserving your edits
        </Card>

        <Card title="Typed variables" icon="list-tree">
          Envelope variables extracted from the input XSD, referenced as workflow parameters
        </Card>

        <Card title="SOAP 1.1 and 1.2" icon="code">
          Version derived from the WSDL, with standard headers set automatically
        </Card>

        <Card title="Testing built in" icon="vial">
          Test operations with value overrides, a curl preview, and full request/response envelopes in run logs
        </Card>
      </CardGroup>
    </Tab>
  </Tabs>

  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **Reusable UI templates**

<Card title="Reusable UI template fixes" icon="puzzle-piece" iconType="duotone">
  Reusable UI templates get more reliable action mapping, undo/redo, and dependency-version handling.

  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **Chat file upload**

<Card title="Attach multiple files in conversational apps" icon="paperclip" iconType="duotone">
  End users can attach multiple files at once in chat-driven apps, and conversational workflows can receive those files as input. Uploaded files are extracted and indexed into a knowledge base scoped to the chat session, so AI nodes answer grounded in that content. Configurable in the Designer and themable, with runtime support on the web renderer.

  Session knowledge bases are isolated per conversation and per tenant, and are excluded from app export.

  Learn more about [chat-driven workflows](/5.9/ai-platform/chat-driven-workflows).
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **File download component**

<Card title="Download files from a rendered app" icon="download" iconType="duotone">
  A new File Download component lets end users download files directly from a rendered app. Configured in the Designer, with runtime support on web and Android.

  For more information, see the [File download](/5.9/docs/building-blocks/ui-designer/ui-component-types/file-download) documentation.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **Conditional event-handler triggers**

<Card title="Fire event handlers based on a condition" icon="code-branch" iconType="duotone">
  Event handlers can now fire conditionally, configured in the Designer and honored by the web and Android renderers.

  For more information, see the [event handlers](/5.9/docs/building-blocks/ui-designer/event-handlers) documentation.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **Image color property**

<Card title="Color control on the IMAGE component" icon="palette" iconType="duotone">
  The IMAGE component now honors a color property, configured in the Designer and rendered on web and Android.

  See the [Image component](/5.9/docs/building-blocks/ui-designer/ui-component-types/image) documentation.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **Save local variable action**

<Card title="Capture a collection row into local data" icon="floppy-disk" iconType="duotone">
  Event handlers gain a **Save Local Variable** action type: on a component inside a collection, it saves the clicked row's item into local data under a configurable key, entirely client-side with no backend roundtrip — useful for capturing the selected row before a follow-up action fires. Not available inside Reusable UI Templates.

  See the [Save Local Variable](/5.9/docs/building-blocks/ui-designer/event-handlers#save-local-variable) documentation.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **Global header and footer**

<Card title="Consistent header and footer across a UI Flow" icon="window-maximize" iconType="duotone">
  UI Flows can now define a **global header and footer** that display consistently across every page of the flow. Each holds content components — containers, forms, buttons, links, custom components — with fixed or static positioning (header fixed by default), so you can build persistent navigation or action bars in one place instead of per page.

  See the [Global Header and Footer](/5.9/docs/building-blocks/ui-designer/ui-component-types/root-components/global-header-footer) documentation.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **Tabular data extraction node**

<Card title="Extract spreadsheet and CSV data into workflows" icon="table" iconType="duotone">
  A new workflow node reads a spreadsheet or CSV file and returns its contents as structured JSON in the node's response key, so the data can drive conditions and data operations downstream. Each row is returned as an object keyed by the file's header row, and for Excel files you can select which sheet to read.

  For more information, see the [Tabular Data Extraction node](/5.9/docs/platform-deep-dive/integrations/integration-designer#tabular-data-extraction-node) documentation.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **Knowledge Base error handling and retry**

<Card title="Recover from failed Knowledge Base ingestion" icon="rotate" iconType="duotone">
  When Knowledge Base ingestion fails, the Knowledge Base page now shows a clear failure status and reason, and a **Retry** button re-runs the last operation after cleaning up partial content so the store is not left inconsistent. The **Update Knowledge Base** workflow node gains a **FAIL** exit handle with per-operation rollback.

  Learn more about [Knowledge Bases](/5.9/docs/platform-deep-dive/integrations/knowledge-base-integration/knowledge-base-overview).
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **AI Developer: generate script-node code**

<Card title="Generate script-node code with AI Developer" icon="sparkles" iconType="duotone">
  In the workflow designer, a script node can call **AI Developer** to generate or refactor its code from a natural-language prompt. Output streams into the test area for a **Keep** or **Discard** review, with edits shown as a diff. Supports JavaScript and Python, and respects existing AI Developer permissions.

  Learn more about [AI Developer](/5.9/ai-platform/pre-built-agents/ai-developer).
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **FlowX Database aggregation**

<Card title="Aggregation pipelines in FlowX Database" icon="layer-group" iconType="duotone">
  The FlowX Database data source adds MongoDB **aggregation pipeline** support, so configurators can group, join, reshape, and compute across documents beyond what a single query allows.

  Learn more about the [FlowX Database](/5.9/docs/platform-deep-dive/integrations/flowx-database) data source.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **AI evaluation experiments**

<Card title="Richer AI evaluation experiments" icon="flask" iconType="duotone">
  <Tabs>
    <Tab title="Multiple datasets">
      An evaluation experiment can run against **multiple datasets** at once (datasets sharing the same AI node), with overall metrics aggregated across them and a dataset column added to the report.
    </Tab>

    <Tab title="Improvement suggestions">
      Experiment reports now show **AI-generated suggestions** for improving the node's prompt, grouped per LLM-as-judge evaluator, beneath the score cards.
    </Tab>
  </Tabs>

  Learn more about [AI evaluations](/5.9/ai-platform/evaluations).
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **AI node output reliability**

<Card title="Schema validation with auto-repair and clearer run errors" icon="shield-check" iconType="duotone">
  <Tabs>
    <Tab title="Validation and repair">
      When an AI node has a **Response Schema**, the response is validated against it and automatically retried with a repair prompt on mismatch. Nodes that still fail return a **categorized error** instead of passing malformed output downstream, and the node's **Logs** show the failure reason, rejected output, the model's reasoning summary, and the retry count.
    </Tab>

    <Tab title="Schema placeholders">
      Response schemas now support `${variable}` placeholders resolved at runtime against workflow data, with substituted values coerced to the schema-declared types — consistent with how other AI-node fields resolve placeholders.
    </Tab>
  </Tabs>

  See [AI node types](/5.9/ai-platform/agent-builder/node-types#structured-output-validation-and-run-diagnostics) for details.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **Template Filling node**

<Card title="Generate documents from templates in workflows" icon="file-signature" iconType="duotone">
  Integration workflows gain a **Template Filling** node, under **Tools** in the node palette, that generates a document from a document template populated with workflow data. Select a template and language, fill the template's variables from workflow data, and choose **PDF** or **DOCX** output. The generated file's ID and download path are returned on the node's response key.

  See the [Template Filling node](/5.9/docs/platform-deep-dive/integrations/integration-designer#template-filling-node) documentation.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **File manipulation nodes**

<Card title="Combine and split files in workflows" icon="file-lines" iconType="duotone">
  Integration workflows add two document nodes, under **Tools** in the node palette: **Combine Files** merges multiple PDFs into one in a given order, and **Split File** breaks a PDF into multiple files by page ranges or one file per page. Both bring document handling that was previously process-only into workflows.

  See the [Combine Files](/5.9/docs/platform-deep-dive/integrations/integration-designer#combine-files-node) and [Split File](/5.9/docs/platform-deep-dive/integrations/integration-designer#split-file-node) documentation.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **DOCX document generation**

<Card title="Generate documents as DOCX" icon="file-word" iconType="duotone">
  Template-based document generation can now produce **DOCX** files in addition to PDF. Set `outputFormat` (`PDF` or `DOCX`; defaults to `PDF`) on the generation request, or pick the format from the new **Output format** dropdown when testing a template in the Designer. Generated DOCX files preserve the template's styles, data, and images. Barcodes apply to PDF output only.

  Learn more in the [Generating documents](/5.9/docs/platform-deep-dive/core-extensions/content-management/documents-plugin/generating-from-html-templates#generating-docx-output) documentation.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **Document Tag (renamed from Document Type)**

<Card title="Document Type is now Document Tag" icon="tag" iconType="duotone">
  The **Document Type** field is renamed **Document Tag** across the surfaces where a document is tagged for storage (Upload File action, Document-Plugin REST responses, AI Document Generation, and the Web Page Extractor). This is a label change; behavior is unchanged.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **Documents tab in database testing**

<Card title="Inspect data while testing database operations" icon="table-list" iconType="duotone">
  The Database Query testing modals gain a **Documents** tab, so you can inspect the database right after running an operation. Multi-record operations report an affected-record count.
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

### **Mistral model provider**

<Card title="Bring your own Mistral API key" icon="key" iconType="duotone">
  **Mistral** joins the model provider catalog as a bring-your-own-key provider. Add it with your own API key under **Organization Settings** → **AI Settings** → **Model Providers**, then assign its models per capability: chat, image understanding, embeddings, and document OCR. The seeded catalog covers the Mistral Large, Medium, Small, Magistral, Codestral, Pixtral, OCR, and Embed model lines.

  Learn more about [AI providers](/5.9/ai-platform/ai-providers).
  <Badge color="green" icon="globe">All Deployments</Badge>
</Card>

***

## Deployment guidelines

FlowX.AI 5.10 is a SaaS-managed release: FlowX handles deployment-level changes such as infrastructure, Kafka topics, and service configuration. For component versions and infrastructure notes, see the [5.10 deployment guidelines](./deployment-guidelines-v5.10). {/* TODO: fill pins via fx-tool at tag-cut */}


## Related topics

- [FlowX.AI 4.7.10 Release Notes](/release-notes/v4.x/v4.7.x-lts/v4.7.10-november-2025/v4.7.10-november-2025.md)
- [FlowX.AI 5.1.10 Release Notes (LTS)](/release-notes/v5.x/v5.1.x-lts/v5.1.10-june-2026.md)
- [FlowX.AI 5.9.0 Release Notes](/release-notes/v5.x/v5.9.0-june-2026/v5.9.0-june-2026.md)
- [FlowX.AI 5.6.0 Release Notes](/release-notes/v5.x/v5.6.0-march-2026/v5.6.0-march-2026.md)
- [FlowX.AI 5.2.0 Release Notes](/release-notes/v5.x/v5.2.0-november-2025/v5.2.0-november-2025.md)