SKILL FILE

Opportunity Tree with AI

Map opportunities for product direction in a simple 3-level tree (outcome to opportunity to bet). Lighter and faster than the full Opportunity Solution Tree. Built for a one-hour planning session.

Download Skill File ↓

What this skill file teaches Claude

Drop one markdown file into your repo. Claude Code learns how to run this entire workflow.

1

Sharp outcome at the top

Forces the outcome to be measurable, specific, and time-bounded before anything else. Fuzzy outcomes produce fuzzy trees.

2

3-7 opportunity branches

User-centred opportunities ("trial users don't see value fast enough") not solution-disguised opportunities ("we need a new onboarding").

3

Evidence per branch

Each opportunity backed by quote count, funnel data, or gut feel (and the source is visible).

4

Bets per opportunity

2-4 specific bets per opportunity branch. Concrete enough to brief tomorrow, loose enough to allow design iteration.

5

T-shirt scoring

Effort, impact, confidence per bet at T-shirt size (XS-XL). Deliberately light. Full RICE handoffs to `/prioritization-engine`.

6

Explicit no's

Lists the opportunities and bets you considered and dropped, with reasons. Stops stakeholder "why didn't you consider X?" questions later.

What you can build with this

Quarterly planning

You have the quarter's outcome. You need 3 bets to take into sprint planning. Build the tree in 1 hour.

Post-research direction-setting

Research has produced 5 opportunities. Map them to a tree, score, pick the top 3 to validate.

Stakeholder alignment

Leadership asks "what could we do about conversion?" Walk in with a tree showing 3 opportunity branches and your top picks.

Pre-PRD divergence check

Before writing the PRD for one solution, build a tree to confirm you considered the alternatives.

Get the full skill file

Everything above is 80% of the skill file. Download the complete version with full implementation details, agent prompts, and ready-to-run scripts.

Common questions

This is the lighter 3-level version: outcome → opportunity → bet. The full OST has 4 levels (outcome → opportunity → solution → experiment), deeper research integration, and runs as a multi-week discovery program. Use this when you need direction by tomorrow. Use full OST (`/opportunity-solution-tree`) when you're running continuous discovery and want the formal Teresa Torres structure.
User-centred and distinct. "Trial users don't see value fast enough" is good (it's a user need, you can validate it). "We should build a new onboarding" is bad (it's a solution disguised as an opportunity). The rule of thumb: if you can imagine 3 different solutions that would address the branch, it's an opportunity. If you can only imagine one, it's a solution.
Cap at 4. Above that the tree becomes a wish-list and the team can't pick. If you have 8 ideas for one opportunity, run an `/ideation-sprint` instead to develop them properly, then return to the tree.
Helpful, not required. With research, opportunities get evidence ratings. Without it, you'll mark opportunities as "gut feel, needs validation" and the recommended next step is to run `/discovery-plan` or `/interview-guide`. The tree still produces value as a structured guess that the team can rally around.
Two reasons. One: it documents what you considered and why you said no, so stakeholders trust the yes's. Two: rejected branches often come back next quarter when context changes. Keeping them in the tree (struck through or in a no list) is cheap and saves rediscovery later.
No. This skill deliberately uses T-shirt sizes (XS-XL) for effort, impact, and confidence. Full RICE / ICE scoring with reasoning lives in `/prioritization-engine`. The opportunity tree is for direction-setting ("which branch are we attacking?"), not detailed roadmap defence. Hand off to prioritisation once you have your top 3 bets shortlisted.

Ready to automate?

SoloStack gives you every skill pre-installed — scraping, marketing, sales, CRM, and more. One repo. Every department.

Book a Call →