System Skills for AI Agents

Your customer-facing AI agents can now take real actions inside Fedna AI — not just talk. They can create tasks, schedule callbacks, hand conversations over to your team, and log private notes.

Available actions

• Create tasks — when a customer requests a callback, a quote, or any follow-up, the agent creates a task assigned to the right team member
• Schedule calls — the agent books an outbound call for a specific time
• Hand over to human — when the conversation needs a human touch, the agent transfers control and assigns the customer to a team member
• Add private notes — the agent logs internal observations that your team can see but the customer cannot

How it works

These actions are built-in — no configuration needed. Your AI agent decides when to use them based on the conversation. When a customer says "Can someone call me back tomorrow?", the agent creates the task and confirms with the customer.

Every action taken by the AI is tracked. Tasks show who created them (the AI agent), calls show they were scheduled by AI, and notes are clearly attributed.

Conversation Actions log

All AI-initiated actions are collected in a dedicated "Conversation Actions" view, so your team can review what the AI has done across all conversations without digging through individual threads.

The bigger picture

Custom Skills (HTTP and Knowledge) let your AI agents connect to external systems. System Skills let them operate within Fedna AI itself. Together, your agents can handle a conversation end-to-end — answer questions from your knowledge base, check availability via API, schedule a follow-up, and hand over to a human — all without a person lifting a finger.