AI agent responses include an action menu for saving findings to your case notes or organizing cited documents into tags. These actions let you capture the agent's work as part of your case record without leaving the conversation.
Saving AI agent responses to case notes
When the AI agent produces analysis worth keeping, click the + button on the response and select Add to Case Notes. The full text of theagent's response gets added as a quoted block in your currently active note, preserving formatting like headings, lists, and emphasis.
This is especially handy for building a running record of key findings during documentreview. Instead of copying and pasting manually, you can capture the agent's analysis in one click and come back to it later when drafting memos, preparing for depositions, or organizing your case strategy.
If you don't have a note open, the response goes into the default note for the case. You can learn more about creating and managing notes in the managing notes guide.
Tagging files cited by the AI agent
When the AI agent's response references specific documents through AI citations, you can tag all those cited files at once. Click the + button and select Tag Cited Files to open a submenu with your existing tags. Pick a tag to add the cited documents, or choose Create Tag to make a new one.
Tags that already contain all of the cited files show a checkmark, so you can tell at a glance which collections already include those documents. This makes it easy to build review sets straight from the agent's research results. For example, after asking the agent to find all documents related to a specific claim, you can tag the results as "Responsive" or "Key Documents" right away, without leaving the conversation.
This action only shows up when the agent's response includes document citations. For general responses that don't reference specific files, the tag option won't appear. You can learn more about organizing documents with tags in the tags documentation.
Redacting keyword matches from AI agent results
When the AI agent runs a keyword search and returns matching excerpts, the action menu includes a Redact Excerpts option. Selecting it opens the batch redaction workflow with the matched content pre-loaded, so you can review each match and apply redactions where needed.
This gives you a direct path from research to redaction: ask the agent to find sensitive content, review the matches in context, then submit the batch. You still review and approve each redaction before it takes effect. The redact option only shows up for responses that contain keyword search results. For more on the redaction workflow, see the redaction documentation.