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

# Athena — Chief of Staff

> The organization's Chief of Staff — a multi-tool operator that acts with the company's voice. Each org can give her any name they want.

## What Athena does

**Athena** is Apollo Space's **Chief of Staff** persona — the default operator that takes action inside the platform. When you need something to **happen** — a lead moved to a new stage, a note attached to a conversation, a quick context lookup to answer a question — Athena is the one who gets it done.

<Info>
  **"Athena" is the canonical name** used in this documentation — but each org
  can **rename** their Chief of Staff under **Settings → Agents**.
  Common choices: "JARVIS," "Tina," "Ada," or the name of a leader
  the persona represents. The behavior is the same — only the name
  shown in chat and prompts changes.

  Throughout this doc we use "Athena" for simplicity. Mentally substitute
  whatever name your org has chosen (visible on the Home screen, at the
  top of the chat, and on the agent card).
</Info>

How Athena differs from the other personas:

| Persona                     | Primary focus                                                             |
| --------------------------- | ------------------------------------------------------------------------- |
| [Scout](/en/agents/scout)   | Pure **research** — finding public information                            |
| [Marcus](/en/agents/marcus) | **Outbound** — writing and sending messages at scale                      |
| **Athena**                  | **Operations** — coordinating tools, executing tasks, maintaining context |

Athena can call on Scout (to research) or Marcus (to send messages) under the hood — she orchestrates when it makes sense.

## What she can do

Athena is multi-tool. She covers the most common day-to-day building blocks:

<CardGroup cols={2}>
  <Card title="CRM" icon="users">
    Create a lead, move stage, attach activity, update custom fields,
    change assignment.
  </Card>

  <Card title="Tasks & boards" icon="list-check">
    Create a task, assign it to a person or agent, change status, post
    an update in the task thread.
  </Card>

  <Card title="Brain" icon="brain">
    Search org documents, read URL content, save a snippet as a
    knowledge capture.
  </Card>

  <Card title="Integrations" icon="puzzle-piece">
    Actions in apps connected via Composio (Notion, Slack, Linear,
    Calendar, and more) — when you have the toolkit enabled.
  </Card>
</CardGroup>

Athena **does not send outbound messages directly** — that is Marcus's role. She can ask Marcus to draft a campaign, but the act of **sending** always goes through the outbound pipeline (with HITL if configured).

## Athena vs. the Digital Twin

Athena is the **organization's** agent, not any individual's. She knows the company context (Brain, org voice) and coordinates work, but speaks with a **neutral, operational voice** — not as any specific person.

<CardGroup cols={2}>
  <Card title="Athena (Chief of Staff)" icon="building">
    The **company's** agent. Runs the org's day-to-day with a neutral voice —
    moves leads, creates tasks, looks up context, coordinates specialists.
  </Card>

  <Card title="Digital Twin" icon="user-astronaut" href="/en/agents/digital-twins">
    A **personal** agent tied to a specific person, spanning all their orgs — speaks
    with the **leader's own voice**. A distinct agent, not a mode of Athena.
  </Card>
</CardGroup>

<Info>
  They **work together**: a leader's [Digital Twin](/en/agents/digital-twins)
  delegates to Athena when the work belongs to the company, and Athena calls
  on the Twin when a communication needs to carry that person's voice.
  One serves the organization; the other represents the individual.
</Info>

## Long-term memory

Unlike Scout (where each conversation is independent) and Marcus (whose memory lives in the shared Brain), Athena has **accumulative memory** within the org's context:

* Learns org preferences over time ("we always respond formally," "this client complained about deadlines in Q1")
* Remembers decisions made in past conversations — no need to re-explain context every time
* Maintains a **continuity journal** across sessions — the morning Athena knows what was discussed the night before

Memory is **isolated per org** (one client's Athena never sees another's context) and **per-user for sensitive data** (personal memories do not leak between members).

## When to use Athena vs. other personas

| Typical request                                                      | Best persona                                        |
| -------------------------------------------------------------------- | --------------------------------------------------- |
| *"Research company X for me"*                                        | Scout                                               |
| *"Send an email to João saying..."*                                  | Marcus                                              |
| *"Create a lead for João Silva, account TechCo, mark as qualified"*  | Athena                                              |
| *"Summarize what happened in this thread + create a follow-up task"* | Athena                                              |
| *"How would I respond to this email?"*                               | [Digital Twin](/en/agents/digital-twins)            |
| *"List the 5 leads who stopped responding last week"*                | Athena                                              |
| *"Send this proposal for the CFO's review"*                          | Athena (with the financial leader's Twin, optional) |

## Limits — where Athena waits for you

Even as the "operator," Athena **does not autonomously execute** actions that involve:

* **Money** — payment confirmation, refunds, financial adjustments
* **High-stakes external communication** — a commercial proposal to a client,
  a public announcement, a regulatory notice
* **Governance decisions** — plan changes, account cancellations,
  member role modifications
* **Destructive actions** — deleting an org, bulk-deleting leads,
  removing an integration

In these cases, Athena **proposes** and waits for human confirmation.

## Next steps

<CardGroup cols={2}>
  <Card title="Digital Twin" icon="user-astronaut" href="/en/agents/digital-twins">
    The leader's personal agent — distinct from Athena, speaks in their voice.
  </Card>

  <Card title="Scout" icon="binoculars" href="/en/agents/scout">
    The persona Athena calls when research is needed.
  </Card>

  <Card title="Marcus" icon="paper-plane" href="/en/agents/marcus">
    The persona Athena calls on to send messages.
  </Card>

  <Card title="Costs" icon="star" href="/en/billing/stars">
    Every Athena call deducts Stars from the org's wallet.
  </Card>
</CardGroup>