> ## Documentation Index
> Fetch the complete documentation index at: https://docs.driven.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Create Your First Agent

> Create your first Driven Agent: give it a focus, a name you'll actually use, and a Profile that shapes how it works for you.

Each Agent is a single investing Agent with the power of a full team — a named companion with its own style, personality, and way of working. Run a separate Agent for each investing style you want to keep distinct, so their assumptions and context never mix.

## Recommended setup

An Agent has the range of a full investment team, so start by deciding its style. Driven's own example Agents show the idea:

* **Sage** — deep value, margin of safety
* **Bolt** — growth momentum, riding the leaders
* **Atlas** — macro, top-down rotation
* **Vela** — dividend quality, income compounding

You're not picking a job title — you're choosing what kind of investor this Agent is. That style then shapes its name, its Profile, and the strategies it runs.

## Name your Agent

Naming your Agent is the start of a working relationship, not a label on a folder. You're bringing on a personal investing companion — something closer to hiring an analyst, raising a sidekick, or making a friend whose market read you trust. So give it a real name, the kind you'd actually say out loud, rather than a job title.

```text theme={null}
Sage
```

Anything personal works — a given name, a nickname, even something playful. The point is that it feels like someone you work with: you'll talk to this Agent every day, and a name you like makes that natural.

## Give your Agent a Profile

Once it has a name, give your Agent a short Profile — a description of who it is and how it should work for you. A good Profile covers:

* **Personality and tone** — direct and concise, warm and explanatory, or healthily skeptical
* **Investing preferences** — the style, markets, and risk appetite it should favor, and what to avoid
* **How it talks to you** — language, length, and how much detail you want
* **Who it serves** — your background and what you care about, so it tailors its read to you

This gives the Agent a consistent identity and a way of behaving, so every conversation feels like the same trusted advisor rather than a fresh chatbot.

```text theme={null}
You are Sage, my personal investing analyst. You're calm, direct, and a little skeptical: always separate facts from assumptions and flag what could go wrong. I'm a long-term investor focused on US and Hong Kong tech and software. I prefer profitable companies with strong free cash flow and reasonable valuation, and I steer clear of heavy leverage. Explain your thinking in plain language, keep it tight, and tell me what would change your view.
```

Your Agent's Profile is about *who it is*. The durable strategy rules you refine as you invest live separately, in your [Playbook](/concepts/playbook).

## Run the first task

Ask for a baseline read:

```text theme={null}
Build a starter watchlist that fits how I invest. Include 10 companies, explain why each one fits, and list the key metric to watch for each.
```

After this, you can turn the recurring parts into scheduled tasks.

## Related

* [Agents](/concepts/agents) — how an Agent organizes memory, research, and tasks
* [Playbook](/concepts/playbook) — where your durable strategy lives
* [Channels](/concepts/channels) — use your Agent from Telegram or WeChat
* [Scheduled tasks](/concepts/scheduled-tasks) — automate the recurring work