This tool is currently in Alpha and has been released for early access and feedback
It may include incomplete features, limited functionality, or be subject to change. Alpha tools are intended for evaluation and validation purposes only and should only be be used in non-production environments.
Adding the tool to an Agent
|
 |
Configuring the tool
Follow these steps to complete the configuration:
- Enter a URL Template
- Enter the full URL of the web page the agent should read.
- You can include dynamic parameters by wrapping them in curly braces (e.g.
{paramater}) - These parameters will be automatically populated by the agent using user input or other contextual information.
- Enter a Description
- Provide a description of what the web page contains or its purpose.
- This helps the agent determine when to use this tool based on the nature of the user’s question.
- Click Save to finalise the configuration.
- Once configured, the agent will be able to dynamically construct the URL, fetch the web page content, and use that information to inform its response.
Considerations
- Limitations – Some web pages are not designed to be scraped and may not expose all content in raw HTML.
- JavaScript dependency – Pages that rely on JavaScript to generate content may appear empty or incomplete when read; in such cases, it may be more appropriate to use an API call to retrieve the data.
- Access restrictions – Ensure the page is publicly accessible or that the agent has valid credentials if the page is behind authentication.
- Content freshness – The retrieved data is a snapshot at the time of reading; for frequently changing content, consider re-reading or using a live API feed.
Related articles:
Configuration