Source Functions

Source functions return object arrays that can be used to dynamically populate dropdown options, or fetch custom fields.

Populate dropdown options

Functions like slack-post-message require the user to provide an ID value for the channel parameter. Source functions help populate such arguments.

For example, when you show the Function UI to the user so they can select a channel, we call the source function slack-list-conversations (using the user's auth credentials) to populate the channel dynamic dropdown.

In fact, when you predict the arguments with Integry AI, we call the same source function to get the list of channels so we can predict the most relevant option.

When you get a function, the description of a field will mention if there is a source function available to list the options for that field.

See line 9 of this sample response for GET /functions/slack-post-message:

{
    "name": "slack-post-message",
    "description": "Post a message in a channel",
    "parameters": {
        "type": "object",
        "properties": {
            "channel": {
                "type": "string",
    
            },
            "attachments": {
                "type": "string",
                "description": "A JSON-based array of structured attachments, presented as a URL-encoded string."
            },
            "blocks": {
                "type": "array",
                "description": "A JSON-based array of structured blocks, presented as a URL-encoded string.",
                "items": {
                    "type": "string"
                }
            },
            "text": {
                "type": "string",
                "description": "The content of the message."
            },
            "as_user": {
                "type": "boolean",
                "description": "(Legacy) Pass true to post the message as the authed user instead of as a bot. Defaults to false. Can only be used by classic apps."
            },
            "icon_emoji": {
                "type": "string",
                "description": "Emoji to use as the icon for this message. Overrides icon_url."
            },
            "icon_url": {
                "type": "string",
                "description": "URL to an image to use as the icon for this message."
            },
            "link_names": {
                "type": "boolean",
                "description": "Find and link user groups. No longer supports linking individual users; use syntax shown in Mentioning Users instead."
            },
            "metadata": {
                "type": "string",
                "description": "JSON object with event_type and event_payload fields, presented as a URL-encoded string. Metadata you post to Slack is accessible to any app or user who is a member of that workspace."
            },
            "mrkdwn": {
                "type": "boolean",
                "description": "Disable Slack markup parsing by setting to false. Enabled by default."
            },
            "parse": {
                "type": "string",
                "description": "Change how messages are treated."
            },
            "reply_broadcast": {
                "type": "boolean",
                "description": "Used in conjunction with thread_ts and indicates whether reply should be made visible to everyone in the channel or conversation. Defaults to false."
            },
            "thread_ts": {
                "type": "string",
                "description": "Provide another message's ts value to make this message a reply. Avoid using a reply's ts value; use its parent instead."
            },
            "unfurl_links": {
                "type": "boolean",
                "description": "Pass true to enable unfurling of primarily text-based content."
            },
            "unfurl_media": {
                "type": "boolean",
                "description": "Pass false to disable unfurling of media content."
            },
            "username": {
                "type": "string",
                "description": "Set your bot's user name."
            }
        },
        "required": [
            "channel",
            "text"
        ]
    }
}

Fetch custom fields

Apps like Hubspot let the user completely customize objects to represent their business. Hence, functions like hubspot-create-contact don't have any typical standard parameters like email, name, etc. Instead, it requires an object parameter called properties in which all the fields of the contact to be created have to be passed. Source functions help populate such arguments.

When you show the Function UI to the user so they can fill the Contact object fields, we call the source function hubspot-list-properties (with objectType: Contact) to fetch all the custom fields for the Contact object.

Similarly, when you predict the arguments with Integry AI, we call the same source function to get the list of fields so we can predict the most relevant values for the fields.

When you get a function, the description of an object field will mention if there is a source function available to list the fields for that object.

Sample response for GET /functions/hubspot-create-contact:

{
    "name": "hubspot-create-contact",
    "description": "Create a contact with the given properties. Call hubspot-list-properties to get the properties.",
    "parameters": {
        "type": "object",
        "properties": {
            "properties": {
                "type": "object",
                
                "properties": {}
            }
        },
        "required": []
    }
}

Call a source function

You can call a source function like any other function via the API or SDK.

In fact, they can typically be called headlessly because they usually don't have any required parameters. You should cache the data to avoid repeated calls.

Good candidates to cache are channel list, user list, issue types, project names, etc. Ultimately, it depends on your use-case, but you can also implement a generic solution for supporting function data.

List all source functions

You cannot list all source functions, per se, because they are context-dependent. Instead, you can list all query-type functions with GET /functions?type=query. To get all query-type functions for just one app, call GET /functions?type=query&app=<app_name>.

See /functions for more options.