Power Platform Community Forum Thread Details

The Get page content tool is supposed to retrieve the HTML page content of a notebook page, but the agent is trying to deserialize it as a json object.

I get the error:

connectorPowerFxErrorEvaluating PowerFx expression resulted in error while invoking the connector : PVA.GetPageContent failed: JsonReaderException '<' is an invalid start of a value. LineNumber: 0 | BytePositionInLine: 0.

Categories:

Calling actions from Copilot Studio

Hello CU06050646-0,

The issue here is a content-type mismatch—the problem isn't with the tool itself. PVA.GetPageContent returns HTML (the first byte is <, indicating <html> or <div>), while the tool binding expects JSON. As a result, the deserializer fails immediately.

Possible solutions:

Update the tool's output schema to string (text/html) and handle the result as Text in Power Fx.
Use a Power Automate flow to call GetPageContent, then return the body with Respond to a PowerApp or flow as a Text output (e.g., PageHtml), and call this flow from the agent. This avoids the JSON binding issue.
For structured data, parse the HTML within the flow and return valid JSON to the agent—avoid parsing HTML in Power Fx.
Set(Topic.PageHtml, Text(GetPageContent.Run(pageId)))

Check the agent's run history to confirm the raw response starts with <, verifying it's HTML before updating the binding.

If you found the information above helpful, I would appreciate it if you could share your feedback.