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

# Your plan

> Two files steer everything: core.md defines what you're building, plan.md tracks what to do about it

Two files form the spine of your project. `core.md` says **what you're building and for whom**; `plan.md` says **where your focus should be right now**. Every skill reads them, and `/whats-next` keeps them honest.

## core.md — the project definition

The single source of truth about your idea. Created during your first `/whats-next` session, it holds:

* **Seed Description** — your original pitch, preserved verbatim and never overwritten. Months later, it's a useful record of where you started.
* **Core** — the structured fields the elaboration conversation pins down: Audience (or ICP for B2B), Problem, Solution, Geography, and for B2B the buyer-vs-user distinction. Fields accumulate over time; a missing one just means "not defined yet."
* **How It Reads** — a two-to-three sentence outsider-facing version of the pitch, written after idea elaboration. If a stranger read only this, would they get it?

When research or interviews sharpen your understanding — say, a much more specific segment emerges — skills will *propose* updating the relevant Core field rather than silently drifting away from it.

## plan.md — focus, steps, and history

```markdown theme={null}
## Current Focus

One sentence: the single most important thing to focus on right now.

## Steps

- [x] Define the idea and target audience
- [ ] **Discover existing competitors**
- [ ] Conduct 5 customer discovery interviews

## Log

### 2026-04-12
What changed in this assessment and why.
```

* **Current Focus** is the at-a-glance answer to "what should I be doing?" — milestone-level, not a task list.
* **Steps** is a cumulative checklist; the bolded item is the current priority. Plans deliberately stay short — the next one or two milestones, not the whole journey.
* **Log** is append-only history: every reassessment adds a dated entry explaining what changed and why, so you can trace how your direction evolved.

## How the plan changes

Only `/whats-next` manages the plan — other skills never edit it. It operates at two speeds:

* **Quick orientation** (the default): reads the plan and your artifacts, checks off what you've completed, and tells you where you stand — the strategic focus plus the single sharpest concrete next move. It never restructures the plan.
* **Full reassessment** (when something meaningful shifted — a milestone completed, results contradicting assumptions, you questioning the direction): your entire project state goes to an independent advisor pass that reasons from the files alone, and its recommendations — new focus, new steps, a log entry — are walked through with you before anything is written.

## Pivots

If foundational fields change — a different audience, a reframed problem, a new solution shape — that's a pivot, and `/whats-next` notices. Instead of leaving stale artifacts around, it walks you through them one at a time: each hypothesis, competitor, and interview script gets a **keep / reframe / archive** proposal that you confirm.

Two reassurances worth knowing:

* **Archiving is reversible.** Archived files stay on disk with a note about why; flipping their status brings them back.
* **Interview evidence is permanent.** Analyses and transcripts are never archived — real customer words remain valid evidence even when the idea around them changes.

## Related

<CardGroup cols={2}>
  <Card title="whats-next skill" icon="compass" href="/skills/whats-next">
    The skill that owns initialization, orientation, and re-planning.
  </Card>

  <Card title="The startup/ workspace" icon="folder-open" href="/concepts/workspace">
    Where these files live and who writes what.
  </Card>
</CardGroup>
