AI Agents in depth, in 65 chapters: creating, instructions, tools, knowledge, behavior, tester, inbox, workflows, analytics and usage, with every setting explained.
An agent with no channels serves nobody. Connect it to your WhatsApp, Instagram, Messenger or website chat from the Connections section, then switch it on so it starts replying.
Learn what an AI agent does in Omnifox, how to give it instructions, connect it to a channel and test it in the sandbox before it talks to real customers.
Learn how to write clear instructions so your AI agent replies with the right tone, limits and actions.
AI Agents essentials in a short video: creating an agent, its instructions, tools, knowledge, channels, the tester, hours and handing over to a person.
Tools turn your AI agent into a teammate that acts: it closes conversations, assigns to a human, updates contacts, books appointments and triggers workflows.
Use the sandbox to chat with your AI agent and validate replies, actions and tone before publishing it.
Set up the handoff from AI to a human agent so no customer gets stuck when personal attention is needed.
Load your business information into the knowledge base so the AI answers with real data instead of improvising.
Understand what tokens are and apply best practices so your AI agent stays useful without overspending.
Set up your AI agent to revive conversations that went quiet and recover customers about to slip away.
Understand in minutes what an AI agent is in Omnifox and the four key concepts you'll configure: model, tone, tools and budget.
In Omnifox a single agent handles chat and calls. Learn when text is enough and when to set up voice in the Voice tab.
Tell activate/pause (replies or not) apart from publish (draft vs live), and learn how to roll out changes with no surprises.
Choose which language your AI agent replies in and decide whether it keeps a fixed tone or mirrors how each customer writes.
Use $ to insert variables (like the contact's name) and ! to reference actions inside your agent's instructions.
Let AI rewrite and improve your agent's instructions, and roll back to any earlier version from the history.
Teach your AI agent to write data into the contact profile, tag contacts and move lifecycle stages, all without human intervention.
Let your AI agent offer real time slots and book appointments on your calendar right from the chat, with no double-booking.
Connect your AI agents to each other and to your automations: hand off to an AI specialist or fire a workflow straight from the conversation.
Connect your AI agent to any external API so it can look up plans, log orders, or call your systems mid-conversation.
How to define the data your agent sends to your API and how to extract just the value it needs from the response.
Use the Test button to run a real call, read the response, and adjust, enable, or delete the action without affecting the conversation.
Learn how to add documents to the Knowledge Base, link them to your AI agent, and order them so it answers with real information from your business.
Set how your AI agent sounds on calls: pick a voice per language, adjust speed, cap reply length, and test the voice before going live.
Set how your AI agent hands a call to a real agent: direct (blind), warm with a whispered summary, or let the agent decide per call.
Tune how creative your agent is and how long each reply can get from the Behavior tab.
Stop the AI from answering message by message: batch the customer's bursts and reply once with full context.
Set up a sequence of automatic reminders when a customer stops replying, choosing AI-generated or fixed text per touch.
Define which days and hours your AI agent answers, and what it does off-shift: a customer message or a transfer to a human team.
Make your agent remember a customer's past chats: how many conversations to summarise and how far back to consider them.
The "Daily token budget" field is still in the agent editor, but it no longer stops the AI: spend is controlled by your plan's monthly allowance and your Omnifox balance.
Two controls to keep a human in the loop: require approval of every reply before it's sent, and auto-handoff to a human when the model isn't confident.
Get the most out of the agent tester: inject a real contact, run actions for real with real mode, inspect the LLM trace, and regenerate or restart the conversation.
See how your AI agent performs: conversations managed, resolved-without-handoff rate, token usage and more.
Omnifox scores your agent's replies and flags weak ones; review them and leave your human feedback.
Compare your agent's current instructions (A) with another version (B) in real conversations and keep the one that works better.
How to write your AI agent's system instructions and a checklist to test them before publishing.
Memory across conversations does work: the agent uses summaries of the same contact's earlier conversations. The separate "Recall past interactions" tool has been retired.
A step-by-step checklist for the most common reasons an AI agent goes silent, ignores a configured tool, or keeps repeating the same reply.
How to turn on the deterministic lock that forces a handoff to a human after your AI agent has replied N times without resolving the case.
How to use Omnifox's curated system prompt template library to set up an AI agent (sales, support, FAQ, receptionist, and more) without writing the prompt from scratch.
The three foundational tools every AI agent in Omnifox has: closing the conversation, handing it off to a human, and leaving an internal note.
How to set up a "receptionist" AI agent that greets every incoming conversation and automatically hands it off to the right specialized AI agent (sales, support, etc.).
Which AI agent tools and token allowance each Omnifox plan includes: AI is on every plan, from Lite to Exclusivo.
How to enable the "Book meeting" tool (book_meeting, Crece plan) so your AI agent sends a Cal.com link; Calendly is coming soon. To book straight into the Omnifox calendar, from Lite.
AI agents answer emails too. How that reply looks, what formatting it keeps, and what limitations it has compared to chat channels.