Updated Agent Workflow UX​

We have the first of a number of updates coming on our AI Workflow builder, with the overall aim to improve the ease of building workflows and to streamline the user experience.

The new workflow builder has a much cleaner UX to simplify the building process, which starts with the nodes themselves:

The code that defines each node is now hidden in the side-menu keeping the main workflow diagram cleaner and easier to understand. Clicking on any individual node will load the relevant code definitions on the right-hand pane.

Managing the node inputs and outputs has been improved. Clicking on the *Manage Inputs and Outputs *button on the side menu launches the input / output management pages.

1. NODE INPUTS Manage inputs that come from other nodes:

2. Easily identity from which nodes variables originate

3. *Required *nodes are clearly marked

4. Quickly convert to user input with the Convert to User Input button

5. Edit specific variables through the new configuration form

1. USER INPUTS Manage inputs that are from users:

2. Easily convert to node input, using the Convert to Node Input button

3. Required nodes clearly marked

4. Edit specific variables through the new configuration form

5. Manage overall input form structure and definition

6. OUTPUTS Manage node outputs:

7. Add / delete variables

8. Edit existing variables through the new Edit Outputs form

9. Visible in the UI variables are clearly marked to easily understand which are returned to users and which are passed through Node Testing

Testing individual nodes has been simplified by adding a Run button prominently on each node in the workflow diagram.

Enabling Run mode allows you to manually test individual nodes by providing the relevant inputs manually and evaluating the node outputs inside of the workflow editor.

Elements of the workflow configuration have been updated to provide much greater controls and options, especially when adding in connector nodes.

To access the configuration settings click on the cog symbol at the top of the page. In this example we have added the Slack tool to the workflow, and the configuration screen shows the actual scopes that have been enabled:

To update any of the specific connector scopes for these tools added to a workflow, click on the configure button. This will allow you to manage the specific access / capabilities for the connectors on a workflow by workflow basis.

More updates in the coming months…

Connector Updates​

A new Azure Storage Container connector has been added that can be authenticated using O-Auth. This can be configured to access a specific container or at account level

The generic O-Auth tool has been developed to make it possible for platform users to be able to create their own connectors to any service that supports O-Auth.

The configuration screen details the parameters that need to be provided to build the connector, as well as any scopes that you would like the connector to have access to:

Please note, the specific configuration parameters and scopes available will be determined by the individual service you are connecting to. Details of how to register and authenticate are often provided as part of providers developer pages.

The Outlook connector tools have been updated to also interact with Outlook calendar, enabling a far wider range of potential use cases

New Models​

OpenAI 5.1 and Gemini 3 have both been added to the platform. To find out how well the new models perform don’t forget to check out our benchmark,

Organization Level GitHub Policies​

A new setting has been enabled to manage organization access to the GitHub connector. To enable this navigate to the Profile menu by clicking on your profile on the top right-hand corner and selecting profile.

On the Profile page then use the menu on the left to select Manage >> Permissions

Config Connectors​

A new option has been added to a number of connectors that allow you to authenticate config connectors to specific instances of these applications (e.g. test or sandbox environments). When providing connector tools to end users these config tools allow you to ensure that users authenticate into this specific instance rather than the standard production environment.

The current connectors that support this setup are:

• Salesforce
• Microsoft Teams
• Outlook
• OneDrive
• Sharepoint

To enable this option when authenticating these connectors, flip the Config only option to on:

Implementing config connectors for users​

To surface these config connectors to end users they can be used in conjunction with the connector tools configured in ChatLLM projects. Under the model configuration page there is a new Config Connectors option.

Using the example below, the Sharepoint connector tool has been added in this configuration. In addition a specific Sharepoint connector has been added into the Config Connectors section.

This means that when a user authenticates the Sharepoint connector in this chatbot, the configuration for that connector will be inherited from the Config Connector that has been specified, rather than the standard domain configuration.

Safety Tool​

We have added a new safety tool feature that supports creating custom prompt evaluations that block or allow messages to the underlying chatbot. This allows deploying organizational defined rules for permissible content within user messages.

Please note this feature is only available for custom chatbots built through a ChatLLM project

Creating a Safety Tool​

Safety Tools can be created in the same way as other custom tools. Navigate to the Abacus home page, then in the left-hand menu select Custom Code >> Tools

Make sure the tool type is Python Function and then build the tool either using the AI assisted approach or directly with python

N.B. When building the tool the input should be a list of messages and the evaluation process should return either {SUCCESS: True} to allow messages or {SUCCESS: False} to block them.

Here is a simple example that looks for a keyword, "abacus", and then blocks the message to the chatbot if it is found.

# get last message
message = messages[-1]["text"]
# Main logic: check presence of 'Abacus' keyword
if "abacus" in message.lower():
    return {'SUCCESS': False} # do not pass message to chatbot
else:
    return {'SUCCESS': True} # pass message to chatbot

Implementing the Tool​

To add the tool, use the new Safety Tool option in the ChatLLM configuration screen:

With the safety tool deployed on a custom chatbot, now each time a user asks a question the tool will evaluate if the prompt is "safe" to pass to the underlying chatbot.

Updated Messaging Connector Bots​

The capabilities of bots deployed into either Slack or Microsoft Teams have been updated to now support chatbots configured with connector tools as part of their configuration. Now results from these tools will be passed through to reflect the same responses you would get from the Abacus UI

Super Assistant​

New Models​

Anthropic's Haiku 4.5 has been added to the platform. To find out how well the new models perform don't forget to check out our benchmark, LiveBench

Abacus AI Deep Agent Apps Permissions​

A new feature has been launched that enables all apps developed through the Abacus AI Deep Agent to be made private. An additional apps permission page gives organizations the option to make all apps private by default, accessing them on a specific organisational DNS and only through user login.

To set this, navigate to the Profile menu in the front-end UI (i.e. the Super Assistant UI):

Then on the left-hand menu select Apps. You will see options for both Organizational and Application level permissions. Setting Organizational Level Permissions to private will ensure all apps are deployed behind a sign-in page.

Organization Level Retention Policies​

A new setting has been enabled to allow configuring retention policies for conversations in Super Assistant. To enable this navigate to the Profile menu by clicking on your profile on the top right-hand corner and selecting profile.

On the Profile page then use the menu on the left to select Manage >> Permissions

Desktop Listener​

As part of our Abacus AI Desktop we have added a new listener functionality, which runs on your machine allowing you to transcribe meetings, tasks or any other audio events.

As the Listener is transcribing you can get real-time answers and insights through the chat feature. Full transcripts can also be exported after recording.

Application User Tools​

Users can now authenticate their own connections into popular applications and perform basic search functionality (e.g. across messages, emails, documents) as well as common actions (e.g. sending an email).

To activate, users can either prompt the system, e.g. "What are the latest emails from Accounting in Outlook" and Abacus will check if the user has authenticated the relevant connector. If not, an OAuth will appear for the user to authenticate, before answering the question:

The other option is to pro-actively authenticate connectors via the First Party Connectors page. This can be accessed through the Profile page (Click on the user profile in the top-right hand corner and select Profile). Once there use the menu on the left to navigate to Connectors >> First Party Connectors.

This page will show you any connectors already authenticated:

To authenticate additional connectors use the Add Connector button, which will pop-up the list of additional applications that can be used

Google Slides Support​

Powerpoint presentations created through Abacus AI Deep Agent can now be exported as a Google Slides doc type, adding to already supported PDF and Microsoft PowerPoint.