Update PA-connector with async CreateRun action

PowerAutomate/DEPLOYMENT.md

@@ -0,0 +1,29 @@
+## Deployment instructions
+
+The deployment pipeline normally takes around 14 days. 
+1. First Microsoft verifies the package (1-2 days)
+2. Testing in a preview environment that Microsoft creates for us
+    a. Make sure you login to the preview account. Username and password will be sent to you in two separate emails
+    b. Select the right environment to test the new connector. It is not the one called "Microsoft (default)", but the other one.
+    c. Start a free trial to give permission to test the new connector since it is a premium feature. [https://learn.microsoft.com/en-us/power-platform/admin/power-automate-licensing/deep-dive-on-specific-license#power-automate-trial-license](docs).
+    d. Test that your connector works in a flow or two.
+3. Press the "Go live" button in the partner center (10-14 days before the connector is rolled out to all regions)
+
+As of 20.02.2026 These are the necessary steps to create the package:
+```
+# Create two solutions in Power Automate. One that contains only the connector, and another that contains the connector and one or more example flows.
+# Export the two solutions, lets call them ConnectorSolution.zip and FlowSolution.zip
+mkdir CradlConnector
+cd CradlConnector
+mkdir PkgAssets
+cp /path/to/readme.md intro.md
+cp /path/to/ConnectorSolution.zip PkgAssets
+cp /path/to/FlowSolution.zip PkgAssets
+zip -r package.zip PkgAssets/              
+zip -r SubmissionPackage.zip intro.md package.zip
+# Upload SubmissionPackage.zip to Azure
+# Rightclick and choose Generate SAS
+# Make sure to set the expiry date to at least 15 days from now 
+# Paste the URL into your Marketplace offer
+```
+

PowerAutomate/Readme.md

@@ -1,75 +1,78 @@
 ## Cradl AI
 
-Cradl AI is a no-code AI platform for automating internal document workflows. Cradl AI enables you to:
+Cradl AI empowers organizations to automate any document workflow with purpose-built AI agents. Fast, flexible and zero code.
+
+🚀️ Create custom AI agents to extract structured data from any document type. <br />
+👍 Route low-confidence AI predictions to a built-in human-in-the-loop validation interface. <br />
+🎉️ Continuously improve your *agent* using feedback from end users. <br />
 
-🚀️ Create customized AI models for any document type in any Latin-based language. <br />
-👍 Deploy a fully fledged *human-in-the-loop* validation UI with one click <br />
-🎉️ Automatically re-train and improve your model based on feedback from end users.<br />
 
 ## Publisher: Cradl AI
 
 ## Prerequisites
 
 A free Cradl AI account. If you don't already have one, you can sign up for free [here](https://cradl.ai/).
 
-## Supported Operations
+## Supported operations
 
 This connector supports the following operations:
 
-### Extract Data From Document
-Extract data from documents like invoices, receipts, order confirmations.
+### Extract data from document
+Extract data from documents like invoices, receipts, order confirmations. 
+Toggle waitForResult to False if you want to run this operation asynchronously and handle the result in another flow.
+Use the trigger "Extracted Data from Document" to handle the result asynchronously in another flow.
 
-### Extracted Data From Document
+### Extracted data from document
 Trigger that runs when data is extracted from a document. 
 
-### Validate Trigger Output
+### Validate trigger output
 Validate output from trigger, to make sure that the output originates from Cradl.
 
-### Get Document 
+### Get document 
 Get the content of a document.
 
-### Get Document Metadata
+### Get document metadata
 Get metadata like name and content-type from a document.
 
-### Create Document (deprecated)
+### Create document (deprecated)
 Create a new document.
 
-### Parse Document with Human-in-the-Loop (deprecated)
+### Parse document with Human-in-the-Loop (deprecated)
 
 Parse a document with *Flows*. This operation runs asynchronous.
 
-### Parse Document (deprecated)
+### Parse document (deprecated)
 
 Parse a document by calling the model directly. This operation runs synchronously.
 
 ## Obtaining Credentials
 
-Log into Cradl AI, and in your *Flow* select either a Power Automate trigger or export and copy `Credentials` to the *Client Credentials* field.
+Log into Cradl AI, and in your *agent* select either a Power Automate trigger or export and copy `Credentials` to the *Client Credentials* field.
 
 ## Getting Started
 
-This quick start guide aims to provide a basic overview how Cradl AI can be integrated in a Power Automate workflow. Please refer to the [official documentation](https://docs.cradl.ai/) for up-to-date documentation.
+This quick start guide aims to provide a basic overview how Cradl AI can be integrated in a Power Automate Flow. Please refer to the [official documentation](https://docs.cradl.ai/) for up-to-date documentation.
 
-#### 1. Set up an Agent 
+### 1. Set up an Agent 
 
-*Cradl AI Agents* enables you to automate internal document processes in a simple, effective and unified way. 
-It reduces the risk of using AI models in production since you can send uncertain documents to a *human-in-the-loop* when necessary. 
+*Cradl AI agents* enables you to automate internal document processes in a simple, effective and unified way. 
+It reduces the risk of using AI agents in production since you can send uncertain documents to a *human-in-the-loop* when necessary. 
 By adjusting *confidence thresholds* of your agent, you can decide when a document should be sent to manual verification and when it's allowed to pass straight through. 
 
-### 2. Configure a Power Automate Trigger
+### 2. Configure a Power Automate trigger
 
-Open the workflow of your Cradl AI Agent, in the *Trigger* section, select *Power Automate* from the list of available integrations. 
+Navigate to your Cradl AI agent, in the *Trigger* section, select *Power Automate* from the list of available integrations. 
 Create a new Flow in Power Automate, and select the action "Extract Data from Document" from Cradl AI. 
-Open the dropdown menu for the parameter *Agent* and choose the Agent you just modified in Cradl AI.
+Open the dropdown menu for the parameter *agent* and choose the agent you just modified in Cradl AI.
 Make sure you have a Trigger in your Power Automate Flow that has a valid file as output.
 
-### 3. Configure a Power Automate Export 
+### 3. Configure a Power Automate export 
 
-Open the workflow of your Cradl AI Agent, in the *Export* section, select *Power Automate* from the list of available integrations. 
+navigate to your Cradl AI agent, in the *Export* section, select *Power Automate* from the list of available integrations. 
 Create a new Flow in Power Automate, and select the trigger "Extracted Data from Document" from Cradl AI. 
 Open the dropdown menu for the parameter *Export Action* and choose the Action you just created in Cradl AI.
 
-## Known Issues and Limitations
+## Known issues and limitations
 
 See [API Limits](https://docs.cradl.ai/reference/quotas).
 
@@ -79,33 +82,14 @@ See [API Limits](https://docs.cradl.ai/reference/quotas).
 
 JPEG, PNG, PDF, WEBP and TIFF.
 
-### How many models do I need?
+### How many `agents` do I need?
 
-One per _document process_. For example, if you want to automate an expense approval process where you process receipts, invoices and airline tickets, we recommend using one model even if you process multiple document types with different layouts.
+One per _document process_. For example, if you want to automate an expense approval process where you process receipts, invoices and airline tickets, we recommend using one `agent` even if you process multiple document types with different layouts.
 
 ### Where is my data stored?
 
 Please refer to our to our [Data Processing Agreement](https://docs.cradl.ai/legal/dpa) and [Privacy Policy](https://docs.cradl.ai/legal/privacy-policy) for more information about how personal data is processed.
 
-## Deployment Instructions
+## Deployment instructions
 
 Refer the documentation [here](https://learn.microsoft.com/en-us/connectors/custom-connectors/paconn-cli) to deploy this connector as a custom connector in Microsoft Power Automate and Power Apps.
-
-As of 17.09.2025 These are the necessary steps: 
-```
-# Create two solutions in Power Automate. One that contains only the connector, and another that contains the connector and one or more example flows.
-# Export the two solutions, lets call them ConnectorSolution.zip and FlowSolution.zip
-mkdir CradlConnector
-cd CradlConnector
-mkdir PkgAssets
-cp /path/to/readme.md intro.md
-cp /path/to/ConnectorSolution.zip PkgAssets
-cp /path/to/FlowSolution.zip PkgAssets
-zip -r package.zip PkgAssets/              
-zip -r SubmissionPackage.zip intro.md package.zip
-# Upload SubmissionPackage.zip to Azure
-# Rightclick and choose Generate SAS
-# Make sure to set the expiry date to at least 15 days from now 
-# Paste the URL into your Marketplace offer
-```
-