> ## 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.

# Planning

> Goals become a board of tasks with owners, priorities get clear, and what's in flight stays visible. Planning as a living thing the Chief of Staff maintains — not a doc that rots.

It's 8am on a Monday. The quarter's goal lives in a pretty slide nobody
has opened since kickoff. Three workstreams moved, two stalled, and you
couldn't say which without asking person by person. The plan exists —
it just became a frozen document, while the real work lives in people's
heads, scattered messages, and a board nobody updates.

With Apollo, planning stops being a file and becomes a living thing the
[Chief of Staff](/en/agents/athena) maintains: she breaks the goal into
tasks, shows what's blocked, proposes the next priority, and flags you
when something slips. **What leaves your plate:** keeping the plan in
sync with reality — chasing status, reshuffling cards, remembering to
nudge.

<Note>
  Planning isn't drawing the perfect plan once. It's keeping the plan
  honest every week. The Chief of Staff does the upkeep; the commitment
  decisions stay yours.
</Note>

## The pattern behind every flow

```mermaid theme={null}
flowchart LR
    A[Trigger<br/>new goal, Monday,<br/>deadline slipping] --> B[Chief of Staff pulls context<br/>from the Company Brain]
    B --> C[Decides + uses tools<br/>boards, tasks, routines, docs]
    C --> D{Commits a deadline,<br/>scope, or budget?}
    D -->|No| E[Organizes and records]
    D -->|Yes| F[Proposes + waits for you]
    E --> G[Becomes memory<br/>the next plan starts smarter]
    F --> G
```

## Four planning flows, concretely

### 1. A goal becomes a board of tasks with owners

You type one sentence into the chat: *"I want to double active pilots
by the end of the quarter."* That's the trigger.

The [Chief of Staff](/en/agents/athena) pulls from the
[Company Brain](/en/features/brain) what already exists — how many
pilots run today, who handles what, what worked in the last batch. On
top of that she **decomposes the goal** into a
[board](/en/features/boards): status columns and a set of concrete
[tasks](/en/features/boards) ("map 20 target accounts", "prep the demo
script", "review onboarding for current pilots"), each with a suggested
owner — a person **or** another agent, like [Scout](/en/agents/scout)
for the research leg.

What she does **not** do alone: declare "this is done by the 30th and
so-and-so owns it." Assigning a firm deadline and committing someone is
a **proposal** — you review the board, adjust owners and dates, and
confirm. From there, the plan exists as trackable work, not a slide.

### 2. A weekly routine that proposes the next priority

Nobody has to remember to "do the weekly planning." The Chief of Staff
schedules a recurring [routine](/en/features/rotinas) for herself —
every Friday afternoon, or Monday morning.

When the routine fires, she sweeps the [boards](/en/features/boards) and
the [CRM](/en/features/crm), cross-references the
[Company Brain](/en/features/brain) (quarter goals, what was agreed),
and assembles a **proposed set of priorities**: what deserves attention
first, what can wait, what's been stuck too long. She records it as a
short [document](/en/features/documentos) and hands it to you in the
chat.

You read it in two minutes and decide. The routine doesn't pick your
priorities for you — it does the tedious sweep and arrives with a
grounded recommendation, for you to simply approve (or change).

### 3. What's blocked or in flight stays visible

This is the quiet work that rots a plan fastest: nobody knows, without
asking, what's actually stalled.

The Chief of Staff cross-references statuses, deadlines, and the task
[update](/en/features/boards) threads. When a task runs past its
deadline, sits without an owner, or stays in the same column for days,
she raises it on her own as a
[proactive notification](/en/features/rotinas) — an item in your inbox,
a bell. It isn't you hunting for the bottleneck; it's the bottleneck
coming to you, with the context for why it matters.

<Info>
  **Why it's worth it, in one line:** what normally only surfaces in the
  status meeting (and sometimes not even there) becomes visible the day
  it stalls — early enough to react.
</Info>

### 4. Re-plan when something slips

A key deadline won't be met. The trigger is the task itself running
late.

The Chief of Staff doesn't just point at the delay. She pulls the
context of dependent tasks, understands what else is affected, and
**proposes a re-sequence**: push milestones, redistribute cards across
owners, flag what should drop from the week. It all arrives as a
proposal on the board, with the reasoning in plain sight.

You approve, adjust, or decline. The re-plan is recorded — and the next
time something similar slips, she already knows how you resolved it last
time, because it became memory.

## Where the agent stops and calls you

<Warning>
  **What the Chief of Staff only proposes — never decides alone:**

  * **Committing the team to a deadline or scope** — the firm date and
    the "who owns it" are yours; the agent suggests and organizes.
  * **Resource or budget decisions** — spinning up a workstream,
    spending on a tool, reallocating people.
  * **Money** — any payment, refund, or financial commitment.
  * **Contract or signature**, **commercial proposal to a client**,
    **regulatory or legal communication**.
  * **Destructive actions** — archiving a whole board, deleting work.

  In all of these the agent prepares the work, shows its reasoning, and
  **waits for you**. The rest — organizing, sequencing, reminding,
  surfacing — it handles.
</Warning>

## Where to start

<Steps>
  <Step title="Give the Chief of Staff a goal, in one sentence">
    In [chat](/en/features/chats), state the objective the way you'd tell
    a colleague: *"I need to organize the March launch."* Let her propose
    the starting board.
  </Step>

  <Step title="Review the proposed board and confirm owners and dates">
    Adjust owners and deadlines — that's yours. Confirming the board is
    the moment the plan turns into trackable work.
  </Step>

  <Step title="Let a weekly routine handle the upkeep">
    Ask for a weekly review [routine](/en/features/rotinas). From there,
    the agent keeps the plan honest — you just decide.
  </Step>
</Steps>

## Next steps

<CardGroup cols={2}>
  <Card title="Boards & tasks" icon="list-check" href="/en/features/boards">
    Where the decomposed goal becomes cards with owner, deadline, and status.
  </Card>

  <Card title="Routines" icon="rotate" href="/en/features/rotinas">
    The weekly review and Monday digest that keep the plan alive.
  </Card>

  <Card title="Chief of Staff (Athena)" icon="user-tie" href="/en/agents/athena">
    The operator who decomposes, sequences, and nudges — in the org's neutral voice.
  </Card>

  <Card title="Management" icon="chart-line" href="/en/use-cases/management">
    The leader's vantage point — the pulse of the operation without micromanaging.
  </Card>

  <Card title="Company Brain" icon="brain" href="/en/features/brain">
    The context the agent pulls goals, agreements, and what already worked from.
  </Card>
</CardGroup>
