Feature Requests

Protect portal with MFA

AI that can identify trends in client's ticket history. Many times "band-aid" fixes are applied to individual support sessions when those symptoms could be indicative of greater issues. Something that automates this process and potentially reports that to the account managers could go a long way in providing long-term resolutions.

I would like a metric to show me the wait time for clients per ticket

From HaloPSA release notes for current beta Version Information Version: 2.204 Application: Halo Release Type: Beta Date Released: 6/08/2025 The Model Context Protocol (MCP) HTTP endpoint allows AI assistants such as OpenAI to connect securely and dynamically invoke tools in the Halo API — such as knowledge search, Ticket logging and updating, or executing runbooks. Example usage in the OpenAI Responses API; POST https://api.openai.com/v1/responses { "model": "gpt-4.1", "input": "Log an Incident to test the Halo Remote MCP Server and assign it to me.", "tools": [ { "type": "mcp", "server_label": "halo", "server_url": "<<yourHaloUrl>>/api/mcp", "require_approval": "never", "headers": { "Authorization": "Bearer <<your-halo-access-token>>" } } ] } This will allow the conversational AI to invoke functions within your Halo API whenever it thinks it needs to use them. In the example above the AI will look at what functions are available in Halo, discover functions for creating a new ticket and assigning it to the agent that the access token is for, and determine that it should use them and execute them one after the other, and respond to the user telling them the ID of their new ticket. Enable the MCP Endpoint from Config > AI, by checking the "Enable the MCP Endpoint" setting. This will then show you some information and the above sample request for using Halo MCP within OpenAI. The tools/functions that are available in the MCP endpoint are the same as for Operations Agents within the Halo Agent Application. These can be customised from the AI config screen in the same way as Operations Agents, by removing the default system tools and opting in to them manually, or adding more Create new Ticket tools with different Ticket Types/Templates, or custom Run Runbook tools. The system tools available here will be added to and enhanced with additional arguments in the future.

I would like to request support for a private end-user message feature, similar to a “whisper,” which would allow sending a message that is visible only to the end user and does not appear in the PSA or in the Thread inbox message history. The primary use case is for ticket closure surveys. Currently, when a ticket is closed, we send survey links via email so they are not included in the ticket notes, preventing the possibility of a technician completing the survey on behalf of the user. If this private message feature were available, we could send a final message in Thread at the time of ticket closure to confirm the ticket is closed and provide the survey link, ensuring the user receives the request even if they miss the email. This would improve survey participation while maintaining ticket integrity and preventing unauthorized survey submissions.

I am looking at the audit trail of a ticket and realize that Thread does not identify who made changes to a ticket. It is important for us to be able to review a ticket to see who may have made changes to the ticket. For example, when I closed a Thread through inbox, It does not show my name on it anywhere.

End users are confused about the OTP emails as the entire solution is white labeled to our brand, but the email comes from thread.

We want to be able to use more SSO providers, but google specifically.

I would like to create an intent that has access to a specific document that adds technical expertise for a specific issue. For example, if I have a document that explains, for the end user, how to release an email from Acronis Quarentine, I would like to attach that document to that specific intent. This would be much easier that try to have the agent troll my database (over two million tickets each with many notes) looking for relevent tickets that might or might not have authoratative information.