Attach chat images to Zendesk Ticketing tickets
The Zendesk Ticketing Create Ticket action can now attach the images that an end user sent to your AI Agent in a chat conversation. The images become files on the ticket. Ada rolls this out for each AI Agent instance.
What’s new
The release adds one variable and these behaviors.
- New variable:
chat_attachmentsis a File Data Token that holds the images from the conversation. To use it, mapchat_attachmentsinto the File Data Token input of the Create Ticket action. - Rebuilt for each ticket: Ada builds the token each time the action runs. It does not include images that the AI Agent withheld as sensitive, images that Ada is still checking, images that Ada flags as malware, or images that Ada cannot read. The token holds about 20 images. If the conversation has more, Ada keeps the most recent images.
- Zendesk Ticketing only: this release does not cover other handoff platforms.
For more information, see Images at handoff.
MCP server: large reads return results in pages
Five MCP server reads now return their results in pages. A large AI Agent configuration no longer arrives as one response that an assistant cannot read in full.
Which reads are paged
These reads are paged by size:
get_ada_configurationget_test_cases, when it lists test casesget_test_runs, for a single run fetched bytest_run_idlist_entitieswithentity_type='variables'list_entities, for a version 2 playbook fetched byentity_id
How to page
Each paged response has next_cursor and total_count. If next_cursor is not null, the assistant passes it back as cursor to get the next page. An assistant reads every page before it answers a question about the whole set. get_test_cases and the variables list also accept limit. The variables list accepts name_filter and scope. With group_by_last_run set to true, get_test_cases now returns one list sorted by last run, newest first, with last_run_at on each case. It no longer groups the cases by date.
Snapshots carry the message when an image is withheld
The Snapshot API now exports and imports the message the AI Agent says when it withholds a sensitive image, in the image_redactions_message section. The section is one value: your sentence, or null when the Agent uses the default sentence.
- Export: the snapshot carries your sentence, or
null. - Import: if the section is present, the import replaces the message on the target Agent.
nullrestores the default sentence. If the section is absent, the message does not change. - Validate: a
mode: validaterun reports whether the import changes the message. - Audit log: if the import changes the message, it writes one Updated row for Image Redactions Message.
For more information, see What a snapshot includes and Redaction settings.