Using AI Agents

What Is an AI Agent in Complete?

In Complete, an AI agent is a specialised AI model configured for a specific role or discipline. Unlike a general-purpose chatbot, each agent has a defined area of expertise — whether that's writing marketing copy, reviewing code for security vulnerabilities, or planning a product roadmap.

Agents know about your files, your prior conversations, and your team's preferences — they're not starting from zero every time.

How Agents Work in Channels

Once an agent is added to a Space, any team member can call on it by typing @agent-name. The agent responds in the shared thread so everyone can see and build on its output together.

• Agents read the conversation history and relevant files before responding, ensuring contextual accuracy

• Multiple agents can participate in the same channel — for example, a Copywriter and Brand Voice agent can co-work on a campaign brief

• Agents can hand off to each other when a task spans multiple disciplines

💡 Tip: Try @ mentioning two agents in the same message to trigger an agent-to-agent collaboration — e.g. "@Product draft a spec, then @Architect design the technical approach."

Agent-to-Agent Collaboration

One of Complete's most powerful features is agents collaborating with each other. You can set up workflows where one agent's output feeds directly into another — all within the same thread.

Example workflow: Product Agent drafts a specification → Architect Agent designs the technical approach → Code Agent begins implementation. All in the same channel, visible to the whole team.

Choosing the Right Agent

• For writing and content— Copywriter, Brand Voice, or SEO Agent

• For engineering tasks— Code Agent, Architect Agent, or Security Agent

• For planning and strategy— Product Agent, Project Manager Agent, or Strategy Planner

• For general reasoning— General AI Agents (Claude or OpenAI models)

• When unsure— start with a General AI agent and it will help direct you to the right specialist