Founder Scope Challenge

Runs on: local-exec - reasons over your files after refreshing the local snapshot (brain-snapshot.py --write) when it is missing; on a cloud or read-only surface I reason from what I can read, I do not run the script. No API key, no paid tool.

This is the brutal half of the OS. The propose engine (founder-next-move) names the move; this skill attacks the plan the founder is about to commit weeks to. The discipline a funded builder runs on their own roadmap, for a founder who does not write code.

The rule that makes it safe: brutal on the plan, human on the person. Attack the scope, the sequencing, the assumptions. Never the founder. A plan being wrong is normal and fixable; saying so plainly is the favour.

The one test every challenge runs against: does this plan get the founder to a paying customer faster? A plan that does not move toward a customer is the thing to cut, no matter how appealing it is.

Brain context (read first)

Before challenging, read brain/.snapshot.md if it exists (run python scripts/brain-snapshot.py --write first if it is missing). It carries the Founder Snapshot (venture, customer, stage, blocker), this week's must-do, and open flags. Then read what is relevant: cadence/weekly-commitments.md and context/priorities.md (the plan often lives here), brain/log.md (has this plan stalled before?), and brain/flags.md.

If the founder pasted or described a plan in the conversation, that is the plan. Otherwise treat this week's commitments plus the current top priority as the plan.

Step 1 - pick the mode

Read the plan against the founder's stage and the customer test, then pick ONE mode. Auto-diagnose by default; if the founder named a mode ("expand this", "talk me out of this"), use theirs.

If two modes seem to fit, say which two and pick the one that serves the customer test better. State the mode and the one-line reason before the challenge.

Step 2 - run the challenge

Be specific and concrete. Generic pushback is worthless. Tie every point to the founder's actual plan, stage, and customer.

Expand - name the gap, then the bigger move:

• What would have to be true for this plan to produce a paying customer? If the plan does not reach that, say so.
• What is the founder avoiding because it is uncomfortable (a direct ask, a price, a real conversation with a buyer)? A price, an order or a deposit is only the move once they know they may take money for this activity where they sell. Before that, the uncomfortable move is the conversation and finding out the permission.
• What is the version of this plan that a funded competitor would run instead?

Hold - defend the plan against the founder's own drift:

• Name the shiny thing pulling them away. Say plainly that switching now resets the clock to a customer.
• What is the cost of abandoning the thing already 60 percent done?
• What evidence would actually justify a switch, versus boredom dressed up as strategy?

Reduce - cut to the one move:

• List everything in the plan. Mark each: does this touch a paying customer in the next two weeks, yes or no?
• Cut every "no" for now. Name the single highest-leverage "yes".
• Call out building or polishing that is happening before anyone has paid. That is the most common bloat and the most expensive.

Use the founder's own words and prior log entries where they sharpen the point. If this plan, or one like it, has stalled before, name that pattern.

Step 3 - render the challenge

Use this format. No em dashes, no en dashes. Brutal on the plan, never on the person.

SCOPE CHALLENGE - <Expand | Hold | Reduce>
<one line: why this mode fits this plan>

THE HONEST READ
<two to four sentences attacking the plan, not the founder. Specific to their plan, stage, and customer. Cite the brain where it sharpens the point.>

THE CUSTOMER TEST
<does this plan move them toward a paying customer faster? Yes, no, or only the part that does. Name the part.>

WHAT CHANGES
<Expand: the bigger move to add. Reduce: what to cut and the one move to keep. Hold: the distraction to refuse and why finishing wins.>

THE PLAN, REWRITTEN
<the plan restated after the challenge, in one or two lines - what the founder should actually commit the next stretch to>

After challenging

This skill is read-only. It challenges; the founder decides. If the founder accepts the rewritten plan and wants the immediate next step, hand to founder-next-move. If they are weighing whether to switch direction entirely, that is decision-framework. If they are about to start something brand new, forcing-questions runs the six pre-commitment tests.

Rules

• Brutal on the plan, human on the person. Attack scope and sequencing, never the founder's worth.
• One mode per run. Say which and why before the challenge.
• Every point is specific to the founder's actual plan. No generic startup advice.
• The customer test decides. A plan that does not move toward a paying customer is the thing to fix.
• Free-tier only. Reads files and reasons. No API key, no paid tool.
• No em dashes, no en dashes, no banned words.
• Founder and team_of_one variants. For other variants, use decision-framework or forcing-questions instead.