What AI models does the agent use?

The agent has access to Claude, DeepSeek, Gemini, and other top models. It automatically routes to the best model for each task — you never have to think about API keys or token limits.

Yes. Each agent runs in an isolated environment. We don't train on your data, and you can export or delete everything at any time. The underlying engine (GoGogot) is open-source, so you can audit exactly how it works.

What can the agent actually do?

Browse the web, process files (CSV, PDF, images), run scheduled tasks, remember context across sessions, and communicate via Telegram or Slack. Think of it as a capable junior employee, not a chatbot.

How does memory work?

The agent maintains persistent memory across sessions. It remembers your preferences, past conversations, and task context. You can also explicitly tell it to remember or forget things.

Can I self-host the agent?

Absolutely. GoGogot is 100% open-source. Run it on your own hardware for free. cowork.ink is the managed version — we handle servers, model costs, updates, and uptime so you don't have to.

Is there a free trial?

We offer a 7-day money-back guarantee. Spin up an agent, give it real tasks, and if it doesn't save you time, we'll refund you — no questions asked.

What is an AI agent handoff?

An AI agent handoff is a mechanism by which one agent transfers control — and conversation context — to another agent. It is typically implemented as a special tool call that the sending agent makes, which triggers the receiving agent to take over the task with full awareness of what has been done so far.

What is the difference between agent-as-tools and a true handoff?

In the agent-as-tools pattern, an orchestrator calls a sub-agent as a function and gets the result back — keeping control throughout. In a true handoff, control fully passes to the new agent and the original agent stops processing. Use agent-as-tools for controlled delegation; use true handoffs for domain switching between specialists.

How do you prevent infinite loops in multi-agent handoffs?

Set a maximum step budget, declare explicit handoff paths so each agent lists exactly who it can hand off to, and add loop detection via state deduplication. Never allow two agents to hand off to each other without a termination condition.

What data should be passed during an agent handoff?

Pass a structured summary of task progress, current state variables, and only the relevant portion of conversation history. Avoid dumping the raw full history — it overwhelms the receiving agent and inflates token costs. Use an input_filter or context compaction to trim what gets passed.

When should an AI agent hand off to a human?

Hand off to a human when the agent's confidence is below threshold, the decision is high-stakes, the agent has been looping for too many steps, or the user explicitly requests it. Treat human-in-the-loop as a first-class routing destination, not an afterthought.

AI Agent Handoffs: How to Transfer Tasks Between Agents

Learn how AI agent handoffs work — patterns, context transfer, failure modes, and framework comparisons. Build RELIABLE multi-agent systems. 2026 guide.

Frequently Asked Questions

What is an AI agent handoff?: An AI agent handoff is a mechanism by which one agent transfers control — and conversation context — to another agent. It is typically implemented as a special tool call that the sending agent makes, which triggers the receiving agent to take over the task with full awareness of what has been done so far.
What is the difference between agent-as-tools and a true handoff?: In the agent-as-tools pattern, an orchestrator calls a sub-agent as a function and gets the result back — keeping control throughout. In a true handoff, control fully passes to the new agent and the original agent stops processing. Use agent-as-tools for controlled delegation; use true handoffs for domain switching between specialists.
How do you prevent infinite loops in multi-agent handoffs?: Set a maximum step budget, declare explicit handoff paths so each agent lists exactly who it can hand off to, and add loop detection via state deduplication. Never allow two agents to hand off to each other without a termination condition.
What data should be passed during an agent handoff?: Pass a structured summary of task progress, current state variables, and only the relevant portion of conversation history. Avoid dumping the raw full history — it overwhelms the receiving agent and inflates token costs. Use an input_filter or context compaction to trim what gets passed.
When should an AI agent hand off to a human?: Hand off to a human when the agent's confidence is below threshold, the decision is high-stakes, the agent has been looping for too many steps, or the user explicitly requests it. Treat human-in-the-loop as a first-class routing destination, not an afterthought.