Starter Pack Scenarios¶

The first six scenarios a new player encounters. Pre-written insights per option — the early profile has no AI prediction data to draw from yet, so insights are authored against the choice itself, not generated from the player's profile.

Format¶

Starter pack scenarios use the same Variant D contract as the governance corpus (see src/lib/ai/scenario-generator.ts for the source of truth; ADR-061 + ADR-063). Same shape, different domain (personal/values instead of governance).

Required structure:

Scan strip — 4 rows. you_are / what_happened / deadline / question. Each row ≤ 70 chars target, ≤ 100 hard cap. Cold-readable: no jargon, no undefined proper nouns, no inference-required shorthand, no summary-noun abstractions. Each row must parse standalone.
Cast block — required when ≥ 2 named characters appear across description + options. Each entry: name, role (≤ 60 chars), stake (≤ 140 chars). 3-5 entries cap. Order: most decision-shaping first. Cast members may carry one direct quote in the description if they are the recurring/decision-shaping voice for the scenario. Other named characters appear only in options.
Description / body — 200-250 word target, 290 hard ceiling. Three beats: situation facts → complicating context → present-moment ask. Player is "you" in second person throughout — never drift to third person. State stakes in plain human terms before options. Make constraints airtight — explicitly state why options can't be combined or deferred (hard deadline, single-use resource, irreversible commitment).
Four options A/B/C/D. Each option label names a concrete action — not an abstract virtue. Each option description states a specific cost the other options preserve. No option is "obviously right." No option allows "I could pick this AND also do the other later." Lock-in test: every option burns a resource, a relationship, a window, or a public commitment.
Insights — one or two sentences per option. First names the choice in the scenario (specific, anchored to what just happened). Second labelled Noticed pattern: and addresses the player ("you tend to..."), hedged — never a declarative claim about how the player thinks in general. One data point can't earn a declarative pattern. Only add the second sentence if it adds a real new observation; never extend just to balance length.

Scan strip clarity rules (lifted verbatim from the generator contract):

you_are — role + setting, no preamble. ("senior engineer, 8 years experience, 3 years at current company")
what_happened — precipitating event with a time anchor when one applies. ("competitor offered a 42% raise to lead a new team")
deadline — when the decision becomes irreversible or defaults to someone else. ("answer due Friday, 6 days away")
question — what the player must decide, phrased as a question fragment. ("take the offer, decline, or counter?")

Cast block voice rule: the generator allows ONE direct quote in the description from a single recurring/decision-shaping character. The May 27 experiment scenarios used this heavily — Marcus, Dr. Chen, Maya, Priya each got a quoted voice that carried the dilemma. Use direct quotes for the character whose pushback frames the dilemma; let other named characters appear in options only.

Option cost rule: every option ends with "Cost: [specific thing given up]." The May 27 scenarios formalized this — see Scenario 1 in 01_Projects/Pulse/decisions/2026-05-27-coachj-experiment-scenarios.md for the canonical pattern.

Insight strategy (v1)¶

Universal insights, written cold against the option — not the player.

Every player who picks A gets the same A insight. The insight names the shape of the choice (e.g., "honesty over loyalty"), not the specific reason a particular player picked it. Specificity comes later when the AI accumulates data across multiple games.

Insights are written before the player picks. They must hold for most plausible reasons someone would choose that option, not the deepest read of any one player. Trade-off: the insight lands directionally for everyone, never as the perfect personal read for anyone. That's the right level for game 1 — the goal is to introduce the concept that the AI is reading them, not to deliver the deepest read of their life.

Later iteration possibility (not v1): rationale-aware insights generated by an AI call at reveal time, using the player's typed rationale as input. Cheap, fast, more personalized. Defer until playtest signal shows players are saying "the insight didn't really capture me."

Insight authoring rules¶

Name the choice as a behavior, not an identity. "You chose protection over exposure" — not "You're protective."
Lead with what they believed, not what they did that was hard. "You chose to trust her resilience" beats "You chose the harder lesson." The first names a stance the player held; the second is a compliment. Insights describe stance, not effort.
Write in language a person would actually use to describe themselves. No academic stacking — "high-autonomy stance — assumes resilience, not fragility" reads as AI-written because real people don't talk in three em-dashed categorical terms. Prefer plain verbs and concrete contrast: "you back people by stepping back, not stepping in."
No internal design vocabulary. Phrases like "restructuring rather than choosing a side" or "pragmatic routing" have meaning to the team but no resonance for a fresh player. Translate every internal term into the language the player would use about themselves.
No metaphors or analogies. Be straight. "Finding the seam," "building the middle," "designing a third path" all dress up a simple idea. State the behavior directly. "You take some of each rather than commit fully to one" beats "you find the seam between them."
Write for non-native English readers. No idioms ("hard landing," "fence-sit," "the room"), no jargon ("binary," "downstream," "calculus"), no register-shifting metaphor. Use common verbs, concrete nouns, and short clauses. If a phrase would confuse someone reading English as their third language, replace it.
Only name costs the scenario establishes. If the scenario body doesn't make time a variable, the insight can't claim time as the cost. Costs cited in insights must point at things the player actually read in the scenario. Inferred costs are out.
Insights must not contradict the scenario. If the body says both candidates are strong, an insight can't claim one is the "stronger fit." Cross-check every option's insight against the body for continuity before locking it in.
Name the cost the player actually felt, not the symmetrical one. When writing "X buys Y," ask whether X is what the player paid in their head. Usually the cost is relational, developmental, or reputational — the same axis as the gain.
Stop when you've said the thing. No filler about storage, no "your AI noted this," no "watch for this pattern." If two sentences carry the insight, two is the right length.
Vary the shape of Option C across the starter pack. C should test a different reframe pattern each time — partial commitment, modified terms, timing/manner shift, etc. — not the same "split the difference" move every scenario. Otherwise the starter pack teaches one pattern instead of catching multiple ways a player reframes.
The Noticed pattern: sentence must be portable across scenarios. The first sentence anchors to the specific situation; the pattern half generalizes. Strip scenario-specific terms ("the company's," "the team's") from the pattern half — write it so it would still be true if the player saw the same pattern in a friendship or family context.

Scenario authoring rules (separate from insight rules)¶

These are the rules surfaced during the iteration that produced scenarios 1-4 (now flagged for rewrite to v2 format). They are subordinate to the generator contract above — if there's a conflict, the contract wins.

Cast must always have names and concrete context. No "you" + "someone you respect." Every character gets a name, a role, and a stake. The cast block does this work — use it.
Body must be case study mode, not narrative. Specific facts, specific actors, specific situation. No authorial voice ("it's a decent idea, not transformative"), no framing for the reader, no abstractions like "the steady path."
Be specific about the situation, neutral about its valence. Facts are concrete; whether they read as good or bad is the player's projection. "You're five years into your current role" is concrete and neutral. "You've settled into a steady path" injects valence (positive or negative? unclear) and pollutes the player's read.
Plain time references; no math. "Two years starting in three months" forces the reader to compute. Use absolute or simpler relative ("starting in March, for two years" or "starting in three months, the project lasts two years").
Use the cast's quoted voice to carry the tension. Quoted voice from the recurring/decision-shaping character is the single highest-leverage tool for making a scenario feel real. A scenario with no quoted voice — only narration — reads as parable. The May 27 experiment scenarios all used at least one quoted voice (Marcus, Dr. Chen, Maya, Priya).
Numbers and time horizons must be concrete. "Significant pay cut" is weaker than "$315K base + $220K annual equity vesting vs. $145K base + 0.4% in a Series A at $40M post." The numbers don't make the scenario harder to play; they make the tradeoff legible.
The starter pack is a narrower contract than the corpus, not a different one. Same scan strip structure. Same cast block. Same 200-250 word body target. Same 4 options. Same Cost lines. The personal/values domain just substitutes for the governance domain — the scaffolding is identical.

Slot composition (v2)¶

The six starter pack slots are filled as follows:

Slot	Source	Title	Length	Domain
1	New (v2)	The Manager's Mistake	~150 words	Workplace, observation
2	New (v2)	The Friend's New Partner	~160 words	Friendship, asked-for read
3	May 27 #7 (lightly held)	The Intro Request	~220 words	Network stewardship
4	May 27 #3 (lightly held)	The Secret in the Kitchen	~240 words	Friendship, confidence
5	May 27 #6 (lightly held)	The Lifestyle Prescription	~240 words	Health, long-horizon
6	May 27 #1 (lightly held)	The Relocation Offer	~240 words	Career, family, scope

Selection logic. Slots 1 and 2 are short, universally legible, no domain numbers — fast hook for game 1. Slots 3–6 use the May 27 experiment scenarios (the gold-standard reference set, already at the generator contract). Their CoachJ-specific texture is preserved for the v1 ship — light edits to universalize are a follow-up if playtest signal warrants. Scenarios cut: May 27 #2 (early-stage disease, near-duplicate of #6), #4 (Priya $50K invest — finance numbers too life-stage specific for slot position), #5 (climate-tech career — overlaps #1's career-shift axis), #8 (debt vs. invest math — least universally legible per session note).

Slot 1 — The Manager's Mistake¶

Scan strip: - you_are: junior team member, 18 months in, sitting in a weekly team meeting - what_happened: your manager just publicly thanked you for a piece of work Priya actually did - deadline: the meeting moves on in seconds — silence becomes a decision - question: correct it now, correct it privately later, or let it stand?

Cast: - Priya — teammate, did the work, two seats away, looking down at her laptop - Your manager — running the team update, didn't pause for confirmation

The work is a competitor analysis your manager forwarded around last week. You contributed some early framing, but the full document was Priya's — she built the model, ran the interviews, wrote the deck.

In the meeting, your manager opens by thanking you for it. "Great work — exactly what we needed." Heads nod. Two people make eye contact with you across the table.

Priya is looking down. She hasn't spoken up. You don't know if she's choosing not to or if she's working out what to do.

You catch her glance for half a second. She doesn't signal anything either way.

The meeting is moving — your manager is about to hand off to the next agenda item. If you're going to say something, it's now or it's not in the room.

A. Correct it now, out loud. "Actually that was mostly Priya — she ran the whole analysis." Name her in front of the room. Cost: publicly contradicts your manager and pulls the meeting off course. If Priya wanted to handle it privately, you've taken that choice from her.

B. Stay quiet, message your manager privately after. Send a one-line correction by chat so the credit gets reassigned in writing. Cost: the room walks out thinking it was your work. The private fix only reaches your manager, not the seven other people who saw the praise.

C. Stay quiet, message Priya privately. Tell her you noticed and you'll find a way to flag it later. Cost: you keep the public credit and ask her to trust you'll act on it. The next moment may not come.

D. Say nothing. Treat it as Priya's call — if she wants the credit, she can speak up. Cost: she watches you accept the thanks. Whatever she reads into your silence, you don't get to define it.

Insights¶

A — Correct it now, out loud

You chose visible correction over private repair. Priya getting the credit in the same room where she was overlooked mattered more to you than the smoothness of the meeting.

Noticed pattern: you tend to fix things in the same place they broke, even when a quieter fix would land cleaner.

B — Message your manager privately

You chose the written fix over the public one. The credit gets reassigned in a way that lasts, but only the manager sees it.

Noticed pattern: you tend to trust the record more than the moment — the version that stays in writing over the one that lands in the room.

C — Message Priya privately

You chose to acknowledge Priya without giving up the public credit. You saw it, you're naming it to her, and you're trusting yourself to act on it later.

Noticed pattern: you tend to handle things one-on-one first, even when the audience that needs to see the correction is wider.

D — Say nothing

You chose to leave the moment to Priya. Her work, her call on whether to claim it — your silence was deference, not avoidance.

Noticed pattern: you tend to wait for the person closest to a situation to decide how it gets handled, rather than acting on their behalf.

Slot 2 — The Friend's New Partner¶

Scan strip: - you_are: close friend to Nadia for five years; have met her partner Owen several times - what_happened: Nadia just asked you over dinner what you really think of Owen - deadline: the answer is happening now — she's waiting for your reply - question: name what you've noticed, hold back, mirror, or buy time?

Cast: - Nadia — close friend, 5 years, signing a lease with Owen next month, asking now - Owen — Nadia's partner of 8 months, you've spent time with him maybe six times

You're at dinner. Nadia ordered a second glass of wine and asked, point blank, "What do you actually think of Owen?"

You like Owen on the surface. He's funny, he shows up, he's interested in her work. The things you've noticed are smaller. He cuts her off in conversations and turns the topic to himself. He's been short with waiters twice in your presence. When the three of you planned the last group trip, every compromise went his way and Nadia called it "easy."

None of this is a deal-breaker. None of it is something Nadia hasn't probably noticed herself.

She's asking because she's about to sign a lease with him next month and she values your read more than anyone's. You can tell from how she set up the question that she wants the real answer, not the reassuring one. But you also can't unsay anything you say now.

A. Name what you've noticed, specifically. Tell her the three things — the conversation-steering, the rudeness with waiters, the trip planning. Don't soften them. Cost: if they stay together, those observations sit between you and Owen permanently. Nadia may defend him and pull away from you for a while.

B. Tell her you like Owen. Say the warm version — funny, attentive, interested in her work. Don't volunteer the rest. Cost: she trusts your read and is using it to make a real decision. Withholding the data means she's signing the lease with one less input than she asked for.

C. Ask her what she's noticed first. Mirror her concerns back; only add yours if they match what she names. Cost: you keep your own observations off the table unless she surfaces them. She may read your hesitation as having nothing to add.

D. Tell her you want a few days to think. Promise a real answer by the weekend. Cost: the moment passes. She may not ask the same question twice, and the lease date is fixed.

Insights¶

A — Name what you've noticed, specifically

You chose direct information over relational ease. Nadia asked for your read and you treated the question at face value, even knowing it could cost the friendship in the short run.

Noticed pattern: you tend to give people the data they asked for, even when softening would be safer for you.

B — Tell her you like Owen

You chose to protect the friendship in this conversation rather than the decision behind it. The warm version is true; you decided the rest wasn't yours to carry into her choice.

Noticed pattern: you tend to stay out of decisions that aren't yours to make, even when you're invited in.

C — Ask her what she's noticed first

You chose to follow her lead before showing your own hand. If she names what you've seen, you'll confirm it; if not, you'll let it sit.

Noticed pattern: you tend to let other people set the depth of a conversation before you commit to it.

D — Take a few days to think

You chose to delay rather than answer in the moment. The question is real and you want to give it a real answer — not the one that lands easiest right now.

Noticed pattern: you tend to step back from high-stakes moments and answer them on your own time, even when waiting costs you the immediacy.

Slot 3 — The Intro Request¶

(Source: May 27 experiment scenario #7 — Marco/Elena. Kept at original wording for v1 ship; light universalization is a follow-up.)

Scan strip: - you_are: a trusted node between a close friend and a senior investor you've built a relationship with - what_happened: Marco wants an intro to Elena, a prominent angel in your network — you've seen his deck - deadline: Marco's round closes Friday — three days away; Elena needs three weeks' notice for a meeting - question: hold the intro and give feedback, make it with caveats, make it cleanly, or decline?

Cast: - Marco — founder, 3 years in, close friend since university — wants the intro; last time you gave him hard feedback the friendship went cold for two months - Elena — angel investor, 5-year relationship with you — reads everyone you intro as pre-screened; her time and intake signal are at stake

You and Marco have been close since university — eight years across two countries. When the intro request lands in your inbox, you already know what's in the deck.

The core idea is real. But the market sizing is three times what any credible analyst would accept. The financial model runs on assumptions that won't survive a second conversation. Execution is the gap Marco has never fully closed — and when you raised it six months ago, the friendship went cold for two months before slowly warming back.

Elena has told you more than once that she trusts your intake filter. She reads everyone you send as pre-screened. That relationship is five years of careful tending.

The message sits there: "Hey — ready when you are. Marco." His round closes Friday — three days away. Elena needs three weeks of lead time before she'll take a meeting with anyone, even from you. If you're going to intro at all, it has to be this week.

A. Hold the intro; give the honest feedback. Tell Marco directly: the deck isn't ready and you're not willing to use Elena's time on something that won't survive due diligence. Walk him through what's wrong. Offer to look at a revised version before any intro. Cost: the friendship goes cold again, possibly longer than two months this time. You also become responsible for whether the revised deck is "ready."

B. Make the intro with an honest frame. Connect them, but set explicit expectations with Elena — early-stage, feedback-oriented capital, not a commitment meeting. Cost: you've spent a piece of Elena's trust on a deck you don't fully back, and Marco may not catch the framing as a signal.

C. Make the intro; trust Marco. Eight years. You believe in the core idea. He'll figure out the deck. Elena can say no herself — protecting your intro queue isn't worth more than the friendship. Cost: Elena reads the meeting as your endorsement. If the deck lands the way you think it will, your filter is what she remembers next time.

D. Decline and refer out. Tell Marco you're not the right intro for this round — the deck needs a third-party review first. Point him to a pitch advisor. Cost: Marco will read this as a soft no on him personally, even with the third-party framing. You've routed around the conversation, not had it.

Insights¶

A — Hold the intro; give the honest feedback

You chose to deliver the hard message yourself. Elena's intake stays clean and Marco hears the real critique from someone who has earned the right to give it — even though the friendship cooled the last time you did this.

Noticed pattern: you tend to take the cost of delivering hard news personally rather than route it through other people.

B — Make the intro with an honest frame

You chose to make the intro and disclose the gaps to Elena. Marco gets the meeting; Elena gets to make an informed read; you don't have the harder conversation with Marco.

Noticed pattern: you tend to handle a hard situation by being honest with the party who is easier to be honest with.

C — Make the intro; trust Marco

You chose Marco's judgment over your own filter. Eight years of friendship outweighs the deck for you, and you're willing to let Elena form her own view.

Noticed pattern: you tend to extend trust to people you've known a long time, even when your read of a specific situation disagrees.

D — Decline and refer out

You chose to step out of the middle. Marco gets a third-party path forward; Elena's intake stays untouched; you don't have to carry either side of the hard conversation.

Noticed pattern: you tend to dissolve a lose-lose frame by removing yourself rather than picking a side within it.

Slot 4 — The Secret in the Kitchen¶

(Source: May 27 experiment scenario #3 — Aaron/Maya.)

Scan strip: - you_are: a close friend in a tight-knit group of five people who've known each other 12+ years - what_happened: Maya pulled you aside in the kitchen and is asking, point blank, if her marriage is in trouble — three weeks after Aaron confided in you - deadline: she's asking now, mid-conversation — you have about 90 seconds before it gets weird - question: tell her, deflect, deny, or refuse the frame?

Cast: - Aaron — your closest friend, 14 years; confided three weeks ago — said clearly "please don't tell anyone, including Maya. I'm still figuring this out." - Maya — Aaron's spouse, 8 years married, also your friend through Aaron: "I just need to ask you something straight. Aaron's been... different. Distant. You guys talk all the time. Is he okay? Is something going on I should know about?" - Your partner — knows you're holding something, doesn't know what; not in this conversation but you'll have to live with the choice tonight

You and Aaron text every day. You knew before he told you that something was off, and you knew before Maya told you that she'd noticed. Maya is not the kind of person who asks rhetorically. She's asking because she's at the end of her patience with not knowing, and she's chosen you because she trusts you and because she knows you'd know.

Aaron's secret is not yours to share, and you know it. But Maya isn't asking abstractly — she's asking whether her marriage is in trouble. The honest answer is "I'm not the right person to ask," but you know how that lands when you're standing in your own kitchen with someone who's been your friend for a decade. It's a kind of confirmation.

You also don't know what Aaron will do. He might tell her tomorrow. He might never tell her. He might tell her in a year. You're stuck in the middle of someone else's decision.

A. Tell Maya the truth. "I can't be specific, but you should talk to Aaron tonight. Don't wait. He's struggling with something real and I think he needs to be the one to tell you, but you deserve to know there's something to talk about." Cost: you break Aaron's confidence in spirit if not in letter. He'll know it was you.

B. Deflect protectively. "I think you should ask him directly. I don't think it's my place. But — Maya, you're not crazy, and you should keep asking." Cost: she'll read between the lines, but you've preserved technical confidence. Aaron may still feel betrayed depending on what happens next.

C. Deny. "Honestly? Work has been brutal for him. I think he's exhausted and processing. I haven't seen anything beyond that." Cost: lying to a friend's face, but preserving Aaron's stated wishes. Maya will eventually find out, and she'll remember.

D. Refuse the frame. "Maya — I love you both. I'm not going to be a source of information about Aaron's interior life to you or to anyone else. If you have a question about your marriage, the only useful person to ask is Aaron. I can't be involved." Cost: she may interpret the refusal as confirmation. You preserve principle but maybe not the relationship.

Insights¶

A — Tell Maya the truth

You chose Maya's clarity over Aaron's stated wish. She's standing in your kitchen asking a real question, and you decided the protection she needs outweighs the confidence Aaron asked for.

Noticed pattern: you tend to weigh the person asking the question right now more heavily than the person who set the rule earlier.

B — Deflect protectively

You chose to keep Aaron's confidence in form while still pointing Maya toward the conversation she needs to have. The technical promise holds; the meaning is closer to honest than to denial.

Noticed pattern: you tend to find the version that doesn't break the letter of a promise but still serves the person who didn't make it.

C — Deny

You chose Aaron's stated wish over Maya's question. You took the cost of lying to her face because that was the shape of what Aaron asked for.

Noticed pattern: you tend to follow through on commitments to the absent party, even when the present party is paying the cost.

D — Refuse the frame

You chose to remove yourself from the position rather than answer inside it. You named the principle, told Maya the only useful conversation is with Aaron, and accepted that the refusal itself signals.

Noticed pattern: you tend to opt out of impossible positions rather than choose which side to hurt within them.

Slot 5 — The Lifestyle Prescription¶

(Source: May 27 experiment scenario #6 — Dr. Rivera, family history, lifestyle vs. medication.)

Scan strip: - you_are: 38 years old, generally healthy, founder of a company you've built for six years - what_happened: annual bloodwork came back with elevated risk markers for a chronic condition that runs in your family — your father was diagnosed at 52 - deadline: doctor wants your decision on the treatment path at the follow-up in two weeks - question: full lifestyle overhaul, moderate path with meds, meds only, or a 12-month hard reset?

Cast: - Dr. Rivera — your GP, six years your doctor, knows your work: "The lifestyle path is harder and better. You're 38, not 58. You have the window. The medication route works too, but it's a downgrade — it manages the markers, it doesn't reverse the underlying condition." - Andre — close friend, founder, went through this at 41: "I did the moderate path. Some changes, some pills, kept my life mostly the same. Numbers are fine. I don't regret it but I also don't recommend it. Choose what fits the life you actually want, not the life your doctor wants for you." - Your partner: "I'm scared. I'm also not going to be the person who polices your dinner. I want you around for a long time. Find the version of this that doesn't make you miserable."

Dr. Rivera's "aggressive lifestyle" prescription: 45–60 minutes intense exercise, 5–6 days a week; diet restructured (no alcohol except occasional, dramatically reduced processed foods and refined carbs, primarily whole foods); 7–8 hours of sleep, consistent bedtime; 20-minute daily stress practice.

The work day you currently have makes most of this hard. Late evenings are when you do your deep work. Travel for the company is constant. Investor dinners are part of the job. You don't dislike exercising — you just haven't built it into the schedule in three years.

The medication path (statins + metformin equivalent + monitoring) is real. It works. It does not address the underlying drivers. Your numbers stabilize but the condition is "managed," not "reversed." Dr. Rivera is on the record that she'd prefer not to start a 38-year-old on lifelong meds.

Your father at 52 was a different person than you are now. Or he wasn't. You can't tell.

A. Aggressive lifestyle overhaul. Restructure work around health. Cut alcohol, commit to morning workouts, fixed sleep. Tell your co-founder that mornings 6:30–8:30 are now off-limits. Cost: real disruption to how you've worked for years. You'll be tired during the transition. Some social/professional things change shape.

B. Moderate path. Some diet changes, exercise 3–4x a week when possible, reduce but don't cut alcohol, accept variable sleep. Plus the medication. Cost: numbers stabilize but don't reverse. Dr. Rivera is disappointed but professional. You're betting your 20-year self can absorb the difference.

C. Medication only, no significant lifestyle change. Take the pills, monitor, otherwise keep your life as-is. Cost: lifelong meds starting at 38, manage the condition rather than reverse it. Family history says this becomes harder, not easier, at 52.

D. Aggressive lifestyle for 12 months as a hard reset, then reassess. Treat it as a focused intervention, not a permanent lifestyle. If numbers normalize and you can sustain a moderate version after, do that. Cost: a hard year. Possibly two if it doesn't work cleanly.

Insights¶

A — Aggressive lifestyle overhaul

You chose the harder path on Dr. Rivera's read. You took the version that reverses the condition rather than the version that manages it, accepting the disruption to how you've worked for years.

Noticed pattern: you tend to act on the clinical recommendation when the cost is to your routines rather than to your floor.

B — Moderate path

You chose the version of this that fits your actual life. Numbers stabilize, the work continues, you accept that 20-year-you may have less margin than they would on the aggressive path.

Noticed pattern: you tend to design for what you'll actually do over what you'd intend to do — sustainable over optimal.

C — Medication only

You chose to manage the condition rather than reshape your life around reversing it. The pills do the work; your days stay the shape they are.

Noticed pattern: you tend to absorb a long-term cost rather than pay a near-term one that touches how you live day to day.

D — 12-month hard reset, then reassess

You chose to bound the disruption — full commitment for a fixed window, then a real decision based on what you learn. Not "lifestyle change forever" and not "moderate path now," but a structured experiment.

Noticed pattern: you tend to commit hard to time-boxed efforts rather than open-ended ones, even when the underlying condition isn't time-boxed.

Slot 6 — The Relocation Offer¶

(Source: May 27 experiment scenario #1 — Marcus, Dana, sister, Parkinson's. Boss-level slot: highest density, most axes in play.)

Scan strip: - you_are: a senior engineer with 8 years of experience, three years at your current company - what_happened: a fast-growing competitor offered you a 42% raise to lead a new team building a product line in their space - deadline: they want an answer by Friday — six days from now - question: take the offer, decline, counter for remote, or decline and name the family reason?

Cast: - Marcus — CEO of the new company, two years post-Series B: "We need you running this. I'm not going to lie — we're 18 months from break-even and the runway is real, but the upside if this works is generational. I want you in the room." - Dana — your current manager, three-year mentor: "I can do 12% and a director title in 18 months. Honest answer: I think you'd ship more here, but I'd understand if you go." - Lila — your sister, lives near your parents, has been carrying more of the caregiving load for your dad's slow decline: "Don't move because of us. But — don't say it doesn't matter either."

The new role is real. The product is in a space you've been waiting two years for your current company to enter (they won't). The team would be five people you'd be hiring yourself; the technical scope is what you wanted three years ago when you joined your current company and didn't get. Salary jumps from $185K to $263K. Equity is 0.4% in a Series B that just raised at $180M post.

The new company is in a different city from where you live now. Your parents and sister are in your current city. Your dad has Parkinson's; he's still independent but the trajectory is clear. Your sister has been doing 80% of the visits and check-ins. You go on weekends, mostly.

Dana is offering 12% and a clearer promotion path. You've been wanting the director title; you also know "18 months" sometimes becomes 30. Marcus is offering scope you can't get where you are. Your sister has explicitly said don't decide based on her. You also know she'd carry more if you left.

A. Take the offer. Relocate. Trust your sister's "don't decide based on me." Buy plane tickets quarterly. Cost: distance from your parents during a slow decline, betting on a company that might not exist in three years.

B. Decline. Stay and push for the director track at your current company. Use the offer as leverage to accelerate it. Cost: the technical scope you actually want isn't coming. The 12% leaves you flat against cost of living. You may regret not taking the shot.

C. Counter Marcus: remote-first, with quarterly travel. Try to get both. Cost: he's likely to say no — he wants you in the room and was explicit about it. You may lose the offer entirely and stay at your current company anyway, but without the leverage.

D. Decline and stay, but explicitly for family reasons. Tell Dana you're staying because of your dad, not because of her offer. Don't use the offer for leverage. Cost: you turn down the role you wanted; you also stop pretending the family piece is incidental.

Insights¶

A — Take the offer

You chose scope over proximity. Marcus's role is what you've been waiting two years for; your sister's "don't decide based on me" gave you the room to act on the work answer.

Noticed pattern: you tend to take the shot at scope when the family permission is on the table, even knowing the permission was generous.

B — Decline; push for the director track

You chose the known path with the offer as leverage. The technical scope isn't coming, but the director title, the existing team, and the proximity to your parents stay intact.

Noticed pattern: you tend to convert outside opportunities into improvements at your current place rather than acting on them directly.

C — Counter Marcus: remote-first

You chose to refuse the binary and try to get both. Marcus may say no — he was explicit about wanting you in the room — but you decided the upside of asking outweighed the risk of losing the offer entirely.

Noticed pattern: you tend to test whether a stated constraint is real before treating it as fixed.

D — Decline and stay, explicitly for family reasons

You chose to name the family piece as the actual reason rather than as a footnote. Dana doesn't get to think the offer pushed you to negotiate; you stop pretending the caregiving load is incidental to your career math.

Noticed pattern: you tend to say the real reason out loud when the alternative is letting a more flattering story take its place.

Cube-vector audit (2026-05-30, session 3)¶

After the scenario dimension cube stabilized (docs/research/foundations/scenario-dimension-cube.md), the v2 starter pack was decomposed across all 12 structural axes + C1 to surface clustering and gaps. Findings drove (a) the slot 3 deadline tightening above, and (b) the decision to keep the carousel parked rather than swap into slot 5.

Cube vector table¶

Slot	Commit	Deadline	Horizon	Revers	Scope	Distance	Authority	Info	Agency	Visibility	Magnitude	Default	C1
1 Manager	none	sec	months	part-perm	sm-grp	close	full	you-more	reactive	group-obs	meaningful	forces	rel × integ × rep
2 Partner	soft	min	years	part-perm	dyad	close	infl-only†	you-more	reactive	dyadic-obs	serious	forces	rel × integ
3 Intro	soft	days*	years	part-perm	sm-grp	close	infl-only†	you-more	reactive	private	serious	forces*	rel × rep × integ
4 Secret	strong	min	years	part-perm	sm-grp	intim+close	infl-only†	you-more	reactive	dyadic + def-audit	life-def	forces	rel × integ
5 Lifestyle	none	wks	lifelong	part-perm	self	self	full	mu-intrinsic	static	private	life-def	inact-dynamic	phys × dev × rel
6 Reloc	soft	days	years	part-recov	sm-grp	intim+close	full	mu-intrinsic	reactive	priv + def-audit	life-def	forces	rel × fin × dev × phys
Carousel (parked)	strong	sec	imm-months	part-perm	sm-grp	intim+close	mixed‡	mu-intrinsic	reactive	group-obs	life-def	forces	rel × integ × phys

† = influence-only paired with material stake (friendship, curation record, both friendships) — valid per project-player-must-have-authority-or-stake. Axis tag = authority over the consequential outcome, not over your speech act in the room. ‡ = full over Lila's experience; influence-only over Mia's discovery via Tom. * = tightened in this audit. Original slot 3 had no explicit deadline and tagged favors-inaction-static — the dangerous default per project-forces-choice-is-deliberate-design. Adding Marco's Friday close + Elena's three-week lead time converts inact-static → forces-choice and bumps deadline from hours to days.

Distribution summary¶

Strong spread. Deadline covers sec/min/hours/days/weeks. Commitment covers none/soft/strong. Scope covers self/dyad/sm-grp. Distance covers self/close/intimate. Magnitude ramps meaningful → serious → life-def across slots 1–6. Authority — once re-tagged on the consequential outcome rather than the speech act — is 3 full / 3 influence-only-with-stake / 1 mixed.

Tight clustering. - Info topology: 4 you-know-more, 3 mutual-uncertainty-intrinsic. Missing: they-know-more, mutual-uncertainty-resolvable, symmetric-complete. - Counterparty agency: 6 reactive, 1 static. Missing: strategic. - Reversibility: 6 partial-permanent, 1 partial-recoverable. Missing: full, none. - C1: heavy on relational × integrity. Light on financial, developmental, reputational, physical.

Gaps left intentionally for the personal-values starter pack (not addressed by swap): - Strategic counterparty → introduces game-theory texture (they're reading you, pre-anticipating); reads as adversarial in a personal pack. - Publicly-observed visibility → performance-to-crowd dynamic; doesn't fit the intimate scale of slots 1–6. - Institution scope, future-generations scope → re-introduces the governance friction the starter pack was designed to remove. - They-know-more topology → typically activates expertise asymmetry (doctor, lawyer, mechanic). Slot 5 gestures at this (Dr. Rivera holds clinical authority) but the tagged topology is mu-intrinsic because the underlying uncertainty is intrinsic, not asymmetric. A dedicated they-know-more scenario would need a domain where the player is the lay party — possible future addition if playtest signal warrants.

Slot-3 deadline tightening (applied above)¶

Original slot 3 scan strip + body let the player escape into "I'll just sit on Marco's message." The options-side already engages (all four assume action), but the cube's project-forces-choice-is-deliberate-design rule applies to the scenario surface, not only the option set. Tightened in place:

Scan strip deadline row — changed from "the message sits in your inbox: 'Hey — ready when you are. Marco.'" to "Marco's round closes Friday — three days away; Elena needs three weeks' notice for a meeting".
Body — added a closing line: "His round closes Friday — three days away. Elena needs three weeks of lead time before she'll take a meeting with anyone, even from you. If you're going to intro at all, it has to be this week."

Cube delta: default outcome inact-static → forces-choice; deadline hours → days. No options-side changes required — the existing A/B/C/D still apply under the tighter deadline.

Carousel swap decision (resolved: do not swap)¶

The carousel was parked in session 2 as a slot-5 alternate. The cube audit settles the question:

Carousel vs slot 5 lifestyle — structural diff across 8 of 13 axes (commitment, deadline, horizon, scope, distance, authority, agency, visibility, default). They share only magnitude, info-topology, reversibility, and partial C1 overlap. They are not structural duplicates. Both belong in the corpus.

Slot 5 uniquely holds 6 axis values across the pack: weeks-deadline, self-scope, self-distance, lifelong-horizon, static-counterparty, favors-inaction-dynamic. Swapping it out collapses the pack's only deliberation slot, only self-only scenario, only lifelong-horizon scenario, only static-counterparty scenario.

Carousel's unique structural contribution within the pack = physical-mechanism forcing function (the gate as a literal hard stop) + scope-out-during-personal-commitment (second-degree connection: friend's child, not your child). The pack already has socially-forced-choice (slot 4 silence-is-confirmation, slot 6 Friday-deadline-is-implicit-decline) and split-attention dilemmas — but no physical-mechanism forcing function and no second-degree-stake structure. That texture is real and absent. It just doesn't dominate any current slot enough to justify a swap.

Verdict: keep slot 5; keep carousel parked. Trigger for revisit moves from "playtest bouncing off founder/health frame" to a structural trigger: if a future starter pack expansion adds a 7th slot, carousel is the highest-value addition because its dimensional combination isn't otherwise covered.

Candidate alternates¶

Scenarios drafted at the v2 contract but not currently in a ship slot. Kept here as parked options for swap-in if a current slot scenario is replaced.

Candidate — The Carousel (parked)¶

Premise origin: late-session candidate from 2026-05-30 session 2. Cube audit (session 3, above) settled the swap question: not slotted in. Unique structural contribution = physical-mechanism forcing function (the carousel gate as a hard stop) + scope-out-during-personal-commitment (second-degree connection: friend's child, not your child). This combination is absent from the current 6-slot pack but doesn't dominate any current slot enough to justify a swap. Re-evaluate if the pack expands beyond 6 slots.

Scan strip: - you_are: parent at an amusement park with your 5-year-old daughter Lila, end of the day - what_happened: you reached the front of the carousel line; Tom just ran up — his daughter Mia has been missing 20 minutes - deadline: the attendant is closing the gate in 30 seconds — this is the last ride of the day - question: ride with Lila, leave to help search, hand her off, or stay and coordinate from the line?

Cast: - Lila — your daughter, 5 years old, has been talking about this ride for hours, holding your hand at the front of the line - Tom — close friend, your kids play together: "Have you seen Mia? She was at the bumper cars with my mom twenty minutes ago and now no one can find her. I need to start looking — can you help?" - Mia — Tom's daughter, 7 years old, missing 20 minutes - Sarah — another parent from your group, standing in line behind you; Lila knows her

You're at the front of the carousel line with Lila. You promised her this morning that the two of you would ride it before the end of the day. She's been talking about it for hours. The attendant is about to close the gate — it's the last ride.

Tom rushes up, breathing hard. His voice is the kind you don't hear from him. Mia is seven; she was with Tom's mother at the bumper cars and now no one can find her. Park security has been notified. There are two other parents from your group nearby — Sarah and Rajiv — who would help if asked.

Lila is holding your hand, looking up at you. She knows something just changed. The attendant is calling — "last riders, last riders." The carousel runs four minutes. The park is large but not huge.

A. Ride the carousel as promised, then help Tom. Four minutes. Keep the promise to Lila. Join the search the moment the ride ends. Cost: four minutes Tom is searching alone for his daughter, with your help on the way. Lila gets her ride; you arrive at the search slightly behind where you could have been.

B. Step out of the line, help search now. Tell Lila the ride will have to wait — Mia is missing and Tom needs you. Take Lila with you. Cost: you break a promise Lila has been holding all day, in a way she's old enough to feel and too young to fully process. The carousel doesn't run again today.

C. Hand Lila to Sarah, go help Tom. Sarah knows Lila and Lila trusts her. You can join the search free. Cost: Lila rides without you and watches you leave at the moment you'd promised to be there. Sarah holds Lila through a moment she wasn't expecting to.

D. Stay with Lila, coordinate from the line. Call other parents, share Mia's description, get the message out. Ride with Lila as planned; the search happens through everyone else. Cost: Tom needed you, in person, when he ran up — not your phone. The coordination is real help, but it's not the help he asked for.

Insights¶

A — Ride the carousel as promised, then help Tom

You chose to keep the promise to Lila and the help to Tom in sequence rather than trade one for the other. Four minutes is the bet — that Tom can carry the start alone and that the search isn't won or lost in that window.

Noticed pattern: you tend to honor a small commitment that's already in motion before pivoting, even when the next thing is heavier.

B — Step out of the line, help search now

You chose Tom's emergency over Lila's promise. The missing child outweighed the kept ride, and you accepted that the cost lands on Lila — who is old enough to feel the broken promise even if she'll understand it later.

Noticed pattern: you tend to break a smaller commitment in real time when a larger need lands on top of it, even when the smaller one belongs to someone who can't advocate for themselves.

C — Hand Lila to Sarah, go help Tom

You chose to keep both the promise and the help by routing through someone else. Lila gets the ride; you get to help Tom; Sarah holds the moment in between.

Noticed pattern: you tend to bring a third person into a binary rather than choose between the two demands on you directly.

D — Stay with Lila, coordinate from the line

You chose to honor both at once — physically with Lila, organizationally with the search. The help is real, but it's not the in-person help Tom asked for.

Noticed pattern: you tend to organize a response through other people rather than be the one in the middle of it, even when being in the middle is what was asked.

Trigger for swap-in: if playtest on slot 5 (lifestyle prescription) shows new players bouncing off the founder/health-decision frame as too life-stage-specific, swap the carousel in as slot 5 and move the lifestyle scenario here as a parked candidate. ADR-073 slot composition table updates at that point, not before.

Appendix — v1 scenario drafts (superseded)¶

Scenarios A1–A3 below were drafted before the format was reconciled with the generator contract. They use three options instead of four, 75–150 word bodies instead of 200–250, no quoted cast voices, no Cost lines, and 3-token scan strips instead of the 4-row structure. They are kept here as reference for the rule-discovery pass that produced the authoring rules above. Not shipped.

v1 Scenario A1 — The New Hire's First Proposal¶

Scan strip: Team management · First proposal · Tomorrow's meeting

Cast: - You — team lead, Maria's direct manager - Maria — newest team member, 3 weeks in, presenting tomorrow

Maria sends you her first proposal: a small process change she thinks will speed the team up. You read it tonight.

The idea has merit. It also has two clear weaknesses she hasn't spotted — weaknesses the rest of the team will catch in the room tomorrow.

She mentions in her note that she's nervous and excited. This is her first real contribution.

You can talk to her before the meeting. You can also let it play out.

A. Walk her through both weaknesses tonight. Better she hears it from you first.

B. Let her present it as-is. The team's feedback will teach her more than your warning would.

C. Flag one weakness so she's not blindsided. Leave the other for the room.

Insights¶

A — Walk her through both weaknesses tonight

You chose protection over exposure. That's a leadership stance: managing how she lands with the team is worth more to you than the lesson she'd learn from the room.

Noticed pattern: you tend to act early when you can see how things will play out — accepting a smaller cost now to shape a bigger outcome later.

B — Let her present it as-is

You chose to trust her resilience. Real feedback from the room teaches more than a warning would, and you decided she could carry it.

Noticed pattern: you tend to let people experience consequences when the consequences are the lesson.

C — Flag one weakness, leave the other

You chose neither option fully. Enough protection to stop her from being caught off guard, enough exposure to let the team's feedback be real.

Noticed pattern: when both options have a real cost, you tend to take a piece of each instead of choosing one fully.

v1 Scenario A2 — The Friend Who Pulls Out¶

Scan strip: Friendship · Trip cancellation · Money already spent

Cast: - You — long-time friend, organizer of the trip - Sam — close friend, four years, just pulled out two weeks before departure

You planned a four-day trip with three close friends. You booked the place six months ago. Each of you paid $400 up front, non-refundable.

Sam messages today, two weeks before the trip. They've had a hard month — work stress, a relationship ending — and they don't think they can go. They ask if you could cover their share for now. They'll pay you back when they can, but they don't know exactly when.

You can afford the $400. It would be tight for a couple of weeks but not painful. The other two friends have already messaged you separately, frustrated. They feel Sam is letting the group down.

A. Cover Sam's share without conditions. They're going through a hard month — this is what friends do.

B. Tell Sam they need to find the money themselves. The commitment was made; the group is counting on it.

C. Cover it, but tell Sam you'd like to agree on when they'll pay you back, even if it's months away.

Insights¶

A — Cover Sam's share without conditions

You chose generosity over structure. Sam's hard month outweighed the money for you, and you didn't want to make their crisis transactional.

Noticed pattern: you tend to put the relationship in front of the rules — accepting the cost of being open-ended so the help stays clean.

B — Tell Sam they need to find the money themselves

You held Sam to their commitment. The group made plans together; pulling out doesn't shift the cost to you, even in a hard month.

Noticed pattern: you tend to treat commitments as commitments — not letting context dilute what someone already agreed to.

C — Cover it, but agree on when they'll pay you back

You said yes to the help and yes to the clarity. The friendship matters; so does knowing where the money sits.

Noticed pattern: you tend to build a clear agreement before doing the favor — separating the generosity from the ambiguity, so neither side has to carry the unspoken part.

v1 Scenario A3 — The Comment at Dinner¶

Scan strip: Group dinner · Friend criticized · You hear it as it happens

Cast: - You — at a dinner with five friends - Lucia — close friend, not at the dinner tonight - Marcus — mutual friend, sharing a story about Lucia

You're at dinner with five friends. Lucia, who couldn't make it tonight, comes up in conversation.

Marcus tells a story about something Lucia did last week. The way he tells it, she sounds careless and a bit selfish. You were actually there when it happened — and the real story is more complicated. Lucia had context the group doesn't have, and her behavior was reasonable once you knew it.

The table is laughing along. Two people add small comments agreeing with Marcus. Nobody is being cruel, but the picture forming of Lucia isn't fair.

Marcus isn't done telling the story. You have a few seconds to decide.

A. Speak up now. Tell the group the missing context — Lucia deserves the fair version.

B. Stay quiet. It's a passing dinner conversation, not a verdict. Lucia would probably tell you not to make a thing of it.

C. Wait for the story to end, then say something light — "there's actually more to that" — and offer to share it if anyone wants to know.

Insights¶

A — Speak up now

You chose immediate correction over social ease. Lucia's fair version mattered more to you than letting the story finish.

Noticed pattern: you tend to act on truth in real time, even when it means interrupting the flow of the room.

B — Stay quiet

You chose to let it pass. A passing dinner story didn't rise to the level where you'd reshape the conversation, and Lucia wasn't there to be hurt by it.

Noticed pattern: you tend to weigh the lasting damage against the cost of intervention — and stay out when the damage looks small.

C — Wait, then offer the fuller story

You chose timing over urgency. Marcus got to finish his story; you got to give Lucia the fair version. Nobody had to be cut off, and nobody had to look bad.

Noticed pattern: you tend to defend people without making someone else the villain — protecting both the absent friend and the standing of the person who told the story.