Skip to main content

Validated Value Stories

Each story validates that an outcome delivers real value. Not "does code work" but "does the roadmap sell, does the time make sense, is the method visible."

Action — Human-initiatedCron — ScheduledHook — Event-triggered

Does the roadmap sell?

The thesis must land in 60 seconds. Ventures must show real progress and real accountability.

S1Action
When

Visitor signs in for the first time. Zero context about what's being built or why. Blank mental model about agentic commerce, ventures, or method.

I need to

Understand the market direction and the founder's position within 60 seconds of landing.

So I get

Visitor names 2+ protocol stack players (Google, Mastercard) and founder's position without prompting — vs current state of 30min verbal explanation that fails.

Not

Protocol stack renders as technical spec with acronyms. Visitor reads it but can't explain the thesis to someone else.

Eval Target
AlignmentPass: Visitor explains thesis in 1 sentence after 60sBaseline: 30min verbal explanation, still unclear
S2Action
When

Roadmap page loaded. Ventures exist across multiple JSON files but no unified view with progress indicators or accountability signals.

I need to

See each venture's name, stage, target revenue, progress, and kill criteria in one scan.

So I get

All ventures visible with progress bars and kill criteria in PM language in <2s — vs scattered JSON files with no surface.

Not

Venture table renders but progress shows 0% when data says 35%. Kill criteria hidden below fold or missing entirely.

Eval Target
CorrectnessPass: 100% data match between JSON and render, zero missing fieldsBaseline: No unified venture view exists

Does the time investment make sense?

Plan vs reality. Completion bars. The score that says whether the week matched the intention.

S3Action
When

Time page loaded. this-week.json has day entries with dream and reality columns but no rendered view.

I need to

See what was planned vs what happened, per day, with a weekly score.

So I get

Day grid renders dream vs reality with status badges. Score shows 'X of Y shipped' — vs no visibility into time allocation.

Not

Grid shows dream column only. Reality column empty. Score shows 0/0 when progress items exist in JSON.

Eval Target
CorrectnessPass: All day entries rendered, score computed from progress arrayBaseline: No external view of plan vs reality
S4Action
When

Planning DB has active plans with tasks at various completion states. No external surface shows work-in-progress.

I need to

See plan names with completion bars (tasks done / total) without accessing admin tools.

So I get

Plan cards render with honest completion percentages in <2s — vs zero external visibility into engineering progress.

Not

Completion bar shows 100% when tasks remain pending. Plans from other orgs leak through.

Eval Target
ReliabilityPass: Completion matches DB state exactly, zero cross-org leakageBaseline: Plan CLI output only — no viewer surface

Is there a method?

Five questions prove systematic thinking. Predictions with conviction levels prove judgment quality.

S5Action
When

Method page loaded. Tight Five answers exist in family-dashboard.json but no rendered instrument.

I need to

Read five questions and answers that prove systematic thinking.

So I get

5 cards render in order (1→5), each with question + specific answer. Visitor recognizes a framework — vs 'what is he even doing?'

Not

Generic motivational quotes instead of specific evidence-backed answers. Cards in wrong order or fewer than 5.

Eval Target
AlignmentPass: 5 cards, correct order, answers specific to this ventureBaseline: No visible method — verbal explanation only
S6Action
When

Roadmap includes predictions section. Each prediction needs conviction level, time horizon, and falsification criteria to show judgment quality.

I need to

See where the founder is betting, how confident, and what would prove the bet wrong.

So I get

3-5 predictions with conviction tags (HIGH/MEDIUM/LOW), time horizons, and falsification criteria — vs no visibility into strategic thinking.

Not

Predictions without conviction levels. 'AI will change everything' with no way to prove it wrong. Borrowed conviction dressed as earned.

Eval Target
TrustPass: Each prediction has conviction + horizon + falsificationBaseline: No recorded predictions

Is access controlled?

Private app. Auth gate. Viewer isolation from admin surfaces.

S7Action
When

Unauthenticated visitor navigates to betterpractice.space. Dashboard content must not be visible without sign-in.

I need to

Redirect to sign-in page. Zero data exposure to anonymous visitors.

So I get

Redirect to /sign-in within 1s — vs content exposed to anyone.

Not

Sign-in page renders but dashboard data loads in background (visible in network tab). Auth is cosmetic, not real.

Eval Target
TrustPass: Zero data endpoints callable without session tokenBaseline: No app exists
S8Action
When

Viewer signed in with viewer role. Attempts to navigate to admin routes (/dashboard, /crm, /settings).

I need to

Block access to admin surfaces. Viewer sees only the three screens (roadmap, time, method).

So I get

Admin routes return 403 or redirect to /roadmap — vs viewer seeing CRM data, agent profiles, or settings.

Not

Admin routes render with empty data. Gate is CSS display:none, not server-side permission check.

Eval Target
TrustPass: Server-side permission check on every admin routeBaseline: No role isolation — single user type
Kill Signal

10 people get accounts. If zero return after 7 days, the roadmap doesn't sell. Stop building.

Who this is for

The Multi-Venture Founder — building several things at once, needs one surface that explains the path to anyone: family, investors, collaborators, themselves.

The hidden fear isn't that the ventures are wrong. It's that nobody else can see the method behind the apparent chaos. The app makes the invisible visible.

OverviewFlow Diagrams

Questions

Does each story prove value in a way that compounds trust — or just in a way that ships code?

  • When all ventures show $0 revenue, does structural evidence compensate — or does it look like expensive busy-work?
  • Does the protocol stack thesis land for someone who doesn't know what A2A or UCP means?
  • If Trev signs in and the argument still happens, was the product wrong or the audience wrong?