tool-foundation-sprint-approach-options
À propos
Cette compétence est un outil d'idéation structuré pour la matinée du Jour 2 d'un Foundation Sprint, forçant la génération de 3 à 7 approches candidates d'une page. Elle prévient le biais de la première idée en imposant la création de multiples options, chacune résumant le "quoi", le "pourquoi" par rapport aux facteurs différenciants, et incluant une représentation visuelle simple. Utilisez-la après la validation du Jour 1 et avant de passer à l'évaluation de l'après-midi avec les "Magic Lenses".
Installation rapide
Claude Code
Recommandénpx skills add product-on-purpose/pm-skills -a claude-code/plugin add https://github.com/product-on-purpose/pm-skillsgit clone https://github.com/product-on-purpose/pm-skills.git ~/.claude/skills/tool-foundation-sprint-approach-optionsCopiez et collez cette commande dans Claude Code pour installer cette compétence
Documentation
Foundation Sprint Approach Options
Day 2 morning of a Foundation Sprint. The team forces itself to generate multiple plausible approaches before committing to one. The skill enforces a minimum of 3 approaches; anchoring on a single approach is the most common Day 2 failure mode.
Family contract: docs/reference/skill-families/foundation-sprint-skills-contract.md. This skill is a member of foundation-sprint-skills.
When to Use
- Day 2 morning of a Foundation Sprint.
- Day 1 is signed; the Mini Manifesto, decision principles, and differentiation chart are committed.
- The team is ready to generate candidate approaches before evaluating them through Magic Lenses in the afternoon.
When NOT to Use
- Day 1 is unresolved. Approach Options without differentiation context produces approaches that miss the strategic position.
- The team has only one approach in mind and is unwilling to generate alternatives. The skill forces minimum 3; if the team refuses, the issue is sprint discipline, not tooling.
- More than 7 approaches are emerging organically. The skill caps generation at 7; beyond that the team is generating features, not strategic approaches.
What This Skill Produces
A single bundled artifact containing:
- 3 to 7 one-page approach summaries, each with:
- Name and label (color, letter, ID)
- One-sentence "what it is"
- "Why it's a good idea" rationale (1 short paragraph)
- Simple doodle or textual visual description
- How the approach serves the two chosen differentiators
- Approach set summary table comparing all approaches at a glance (label, name, capture mechanism, recall mechanism, primary trade-off).
See references/TEMPLATE.md for the canonical structure and references/EXAMPLE.md for the Brainshelf example.
Sequence (75 minutes)
Step 1: Frame the approach space (5 min)
The facilitator restates the differentiation: "We are committing to [differentiator 1] and [differentiator 2]. Approaches that fail either of these are out of scope." This sets the boundary; the team generates inside it, not outside it.
Step 2: Silent ideation (15-20 min)
Each team member generates 2-4 candidate approaches silently. Cluster duplicates. Surface 8-12 candidates.
Step 3: Cluster and select 3-7 (15-25 min via Decider call)
The team clusters similar approaches and the Decider narrows to 3-7 candidates that will be one-page-summarized. The 3-minimum is enforced; if the team produces 2 candidates after clustering, push the team back to ideation to generate at least one more.
Step 4: One-page summarize each approach (25-40 min)
Each team member takes 1-2 approaches and writes the one-page summary. Each summary names:
- Label: a color, letter, or short identifier (Yellow, Blue, Approach A, etc.)
- What it is: one sentence the customer would understand
- Why it's a good idea: short paragraph naming the customer value and the team's ability to deliver
- Visual: a simple doodle or textual description of what the customer sees or does
- How it serves the differentiators: one or two lines per chosen differentiator
Step 5: Cross-summary review (5-10 min)
The team reviews the full set, flagging summaries that don't fit the differentiators or that overlap so heavily with another approach that they're duplicates. The Decider approves the set advancing to Magic Lenses.
Approach Generation Discipline
The skill enforces five rules at decision-point:
- Minimum 3, maximum 7. Fewer than 3 means the team is anchored on one idea; more than 7 means the team is generating features.
- Each approach must be a strategic path, not a feature. "Add a settings screen" is a feature; "make capture the home screen" is a strategic path.
- Each approach must serve both chosen differentiators (not just one). An approach that wins on differentiator 1 but fails differentiator 2 should be either revised or dropped.
- Each approach must be visually describable. If the team can't draw it on a card, the approach is too abstract for the sprint.
- No first-idea bias. The first approach the team thought of should be included only if it survives the differentiation check; many teams find their initial idea is not the strongest after generating alternatives.
Inference Inputs
| Input | What the skill does with it |
|---|---|
| Basics bundled artifact | Reads target customer to ensure approaches are designed for them, not for an adjacent customer |
| Differentiation bundled artifact | Reads the 2 chosen differentiators and the 2x2 position; flags approaches that miss either differentiator |
| Approach candidates (optional) | If pre-supplied, pre-populates the silent ideation board; team adds and refines rather than starting cold |
Common Pitfalls
- Generating features instead of approaches. "Add notifications" is not an approach; it's a feature. The skill enforces strategic-path framing.
- Too few options. Stopping at 2 approaches because "those are obviously the choices" anchors the team. Force a third even if it's intentionally weaker; it surfaces the trade-offs.
- Approaches that fail one differentiator. An approach that beats differentiator 1 but loses differentiator 2 has rejected the Day 1 strategic commitment. Either drop it or revise it.
- Skipping the visual. "I can describe it in prose" defeats the purpose. The visual forces concreteness.
- Overlap masquerading as distinct approaches. Two summaries that differ only in implementation detail are one approach. Cluster them.
Decider Role
The Decider's job during Approach Options:
- Restate the differentiation boundary at the start.
- Narrow the clustered candidates to 3-7 for one-page summary.
- Approve the final set advancing to Magic Lenses; reject summaries that drift from the differentiation.
The Decider does NOT pick a top approach in this skill. Magic Lenses produces the top bet; Approach Options produces the candidates Magic Lenses will evaluate.
Canonical Sources
- Character Capital. "Foundation Sprint guide." Approach generation agenda.
- Knapp, J., and Zeratsky, J. Click. Day 2 morning sequence.
Cross-Skill Usage
Prerequisites: tool-foundation-sprint-differentiation. The Day 1 strategic position is the load-bearing input.
Next invocation: tool-foundation-sprint-magic-lenses in the afternoon. The approach set produced here is the input to Magic Lenses scoring.
Decider Checkpoint
This skill ends with a Decider Checkpoint in references/TEMPLATE.md. The Decider signs off on the set of approaches advancing to Magic Lenses, confirming none are out-of-scope and none are duplicates. Without sign-off, Magic Lenses begins with an unstable candidate set.
Dépôt GitHub
Compétences associées
content-collections
MétaCette compétence propose une configuration éprouvée en production pour Content Collections, un outil axé sur TypeScript qui transforme des fichiers Markdown/MDX en collections de données typées de manière sûre avec une validation Zod. Utilisez-la lors de la création de blogs, de sites de documentation ou d'applications Vite + React riches en contenu pour garantir la sécurité de typage et la validation automatique du contenu. Elle couvre tout, de la configuration du plugin Vite et de la compilation MDX à l'optimisation des déploiements et la validation des schémas.
polymarket
MétaCette compétence permet aux développeurs de créer des applications avec la plateforme de marchés prédictifs Polymarket, incluant l'intégration d'API pour le trading et les données de marché. Elle fournit également une diffusion de données en temps réel via WebSocket pour surveiller les transactions en direct et l'activité du marché. Utilisez-la pour mettre en œuvre des stratégies de trading ou pour créer des outils traitant les mises à jour de marché en direct.
creating-opencode-plugins
MétaCette compétence aide les développeurs à créer des plugins OpenCode qui s'interconnectent avec plus de 25 types d'événements tels que les commandes, les fichiers et les opérations LSP. Elle fournit la structure du plugin, les spécifications de l'API événementielle et les modèles d'implémentation pour les modules JavaScript/TypeScript. Utilisez-la lorsque vous avez besoin d'intercepter, de surveiller ou d'étendre le cycle de vie de l'assistant IA OpenCode avec une logique personnalisée pilotée par les événements.
sglang
MétaSGLang est un framework de service LLM haute performance spécialisé dans la génération rapide et structurée pour les workflows JSON, regex et agentiques grâce à son cache de préfixe RadixAttention. Il offre une inférence nettement plus rapide, particulièrement pour les tâches avec des préfixes répétés, ce qui le rend idéal pour les sorties complexes et structurées ainsi que les conversations multi-tours. Choisissez SGLang plutôt que des alternatives comme vLLM lorsque vous avez besoin d'un décodage contraint ou que vous construisez des applications avec un partage étendu de préfixes.
