Review and refine AI-generated content using natural conversation - just tell your agent what to change.
Every agent can receive and analyze images and documents. This is especially useful for:
Click Paperclip
Select files from your device
Drag & Drop
Drop files anywhere in the chat
Paste Images
Ctrl/Cmd+V to paste screenshots
Supported Formats
Images (PNG, JPG, GIF, WebP), PDFs, Text files, CSV, JSON, Word documents (.docx), Excel spreadsheets (.xlsx)
Agent Creates Approval
Content awaits review
You Chat with Agent
Voice, text, or images
Agent Modifies Content
Based on your feedback
You Approve
Workflow resumes
“Do I have any pending approvals?”
Yes, you have 1 pending approval for “Agent Showcase Copy”. The headline is “Streamline Your Property Management” and the description talks about AI-powered efficiency. Would you like me to read the full details?
“Make the headline more exciting”
Done! I've updated the headline to “Revolutionize Your Property Empire with AI”. Want me to make any other changes, or should I approve it?
“Approve it”
Approved! Your agent showcase is now live. The workflow is continuing to the next step.
“Show my pending approvals”
“What needs my attention?”
“Read me the details”
“Change the headline to...”
“Make it sound more professional”
“Shorten the description”
“Approve it”
“Reject this - it's not ready”
“Looks good, let's publish”
“Go back to step 2”
“Show me the change history”
“Undo that change”
Every agent automatically has access to these HITL tools through the unified chat interface:
Microphone Button
Click to start/stop voice input
Speaker Button
Toggle AI voice responses
Browser Support
Voice features use the Web Speech API. Works best in Chrome, Edge, and Safari. Firefox has limited support.