Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Many organizations store important operational and business data in SharePoint lists, such as help desk tickets, course catalogs, backlogs, and more. After you connect SharePoint lists to your Copilot Studio agent, it can answer questions grounded in structured SharePoint data.
Note
Features in this article are powered by the GitHub Copilot harness, which uses Copilot Credits for usage-based billing. Learn more about Copilot Credits in Overview of billing for agents powered by the GitHub Copilot harness.
Usage-based billing applies to using, building, testing, and evaluating agents. These actions might consume Copilot Credits.
This functionality supports the following capabilities:
Natural language lookup and FAQ experiences with semantic understanding capabilities rather than an exact match
Analytical and aggregation queries over structured list data
Large list scalability for early enterprise scenarios
Analytical queries over structured data
Agents can perform analytical reasoning over SharePoint list data.
Example queries
"What is the average lead time to close a ticket?"
"How many procurement requests were submitted for MacBooks?"
"Which category has the highest number of open requests?"
Enterprise scale list support
This preview supports large scale lists. The supported scale is:
| Capability | Public preview support |
|---|---|
| Maximum supported list size | Up to 120,000 rows. You can connect larger lists, but quality and latency might degrade after 120,000 rows. The best performance occurs when total rows are limited to 120,000 rows across all lists. For example, 10 lists of 12,000 rows, or 3 lists with 35,000 rows. |
| Lists per agent | Up to 10 lists. |
| Models | GPT 5.4 and above, and Sonnet models. Results are best with GPT 5.4 and above, and Sonnet models. The other models are still supported, but the quality of the returned results isn't as high. |
Note
Questions requiring analysis of all 10 lists and each of the 35,000 rows might get throttled or subjected to high latency.
Best practices
To retrieve the best results from SharePoint lists, follow these practices:
Use clean and structured list data with clear column names.
Name your list appropriately when you add it to an agent.
Use supported column types, and use descriptive column names.
Keep related data within the same list where possible.
Run evaluations and validate your queries before production rollout.
Add a SharePoint list
Open your Copilot Studio agent.
Go to the Build tab.
In the components panel, select the Knowledge button.
In the Add knowledge dialog, select SharePoint.
Select Browse items to find the list you want, or enter the URL of a specific SharePoint list.
When you select Browse items, the dialog is automatically populated with My Lists and Recent Lists. My Lists only displays lists created in the SharePoint Lists app. All other lists appear in the Recent Lists. If you don't see a list you need, open SharePoint and access the list, and the list then appears in Recent Lists.
You can select up to 10 lists at a time. For the best results, only use up to 10 lists per agent. Lists that contain more than 35,000 rows affect quality and latency.
Each list is added to the agent's knowledge sources as individual items. To add more lists to the agent, use these steps again and select more lists.
Note
If you want to use a SharePoint list with shared access but it isn't displayed in the Recent Lists, paste the URL for the list into the URL selection.
Add a name and a description. Provide a detailed description, since the description aids the agent orchestration and helps the agent understand the context of the knowledge source.
Select Add to agent to finish adding the knowledge source.
After adding the SharePoint list, test your agent to verify the expected results are returned.