An agent has 3 macro-settings that you can edit:
Context & Instructions: Here you can provide your agent with context and instructions in plain text, helping it understand its objectives and which brand/product it represents. You can write in all the languages you want
Interaction: These settings determine how the agent should interact with users
Moderation: These settings control how the agent should moderate incoming comments
The key thing to know about an agent is that by default, it will generate AI responses to received comments but won't post them automatically. If you want the agent to operate autonomously, you need to enable the "Post Replies Automatically" option.
Context & Instructions
This section enables providing additional context to improve your AI agent's understanding and responses. When ReplyZen receives a comment, the managing agent:
Automatically analyzes the post content where the comment was made (image/video, copy, links, and previous comments)
Uses information from the knowledge base section
There's also an option to add additional context to provide extra instructions guiding agent behavior. This section allows you to input such instructions using natural language.
You can provide specific information in all the languages you want.
"Interaction" settings
The Interaction section allows you to configure how your agent interacts and communicates.
Enable Replies with AI: Allow the agent to generate responses to comments
Respond on (Facebook Only - technical limitation due to official API restrictions that allow certain actions only on Facebook and not on Instagram): Select where to respond (Every post / Only ads / Only organic)
Post Replies Automatically: Publish responses without manual approval
Like Comments Automatically (Facebook Only - technical limitation due to official API restrictions that allow certain actions only on Facebook and not on Instagram): Enable automatic likes on each comment
Tone of Voice: Set agent's response tone
Use Emojis: Enable/disable emojis in responses
Response Time:
Natural: Simulate human response times
As soon as possible: Immediate response
Language: Manual selection or automatic detection (ReplyZen tries to understand interaction's language from post and comments texts)
Answer Length: Concise, Standard, or Detailed
Append Link (Facebook Only - technical limitation due to official API restrictions that allow certain actions only on Facebook and not on Instagram):
Auto: Automatically insert relevant links, based on post context and historical data
Manual: User-defined link. You can use this option if you want your agent to always reply with a specific link
No: No links
"Moderation" settings
The Moderation section manages all settings related to comment moderation and filtering:
Enable Moderation: Activate automatic moderation
Moderate on (Facebook Only - technical limitation due to official API restrictions that allow certain actions only on Facebook and not on Instagram): Select which posts to moderate (Every post / Only ads / Only organic)
Hide All Comments: Automatically hide all comments. You can always unhide them from the "Comments" section
Hide Comments with Attachments (Facebook Only): Hide comments with media/links
Custom Instructions: Custom moderation rules. Here you can manually add in plain text custom instructions about moderation. E.g: "Hide comments that talk about politics".
Note on Facebook Features
Features marked with "Facebook Only" are exclusively available for the Facebook platform due to Meta's API restrictions for Instagram.
Aside from its settings, an agent also has a "Knowledge" section. Learn more about it here: Knowledge management