Guide · Agent skill
How to use the Opportunity Solution Tree agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The opportunity-solution-tree skill teaches an AI agent to build an Opportunity Solution Tree, a visual framework from Teresa Torres' Continuous Discovery Habits, for a product team. It is for product managers, product trios and students learning discovery. After this guide you will know what the skill does, how to install it, how to ask for a tree with a plain request, and what to check in the output.
Opportunity Solution Treephuryn/pm-skills · MIT Download the skill Quick start
- 1Download the ZIP. Download the skill ZIP from this page. It contains SKILL.md.
- 2Unzip into a skills folder. Unzip to ~/.claude/skills/ for personal use or to a project's .claude/skills/ folder for one project.
- 3Or upload in the Claude apps. Alternatively, go to Settings → Capabilities → Skills in the Claude apps and upload the skill.
- 4State your desired outcome. Name one measurable outcome you want to improve, such as a retention or activation metric.
- 5Provide your research. Paste interview notes, survey answers, analytics or feedback so the agent can identify opportunities.
- 6Review the tree and experiments. Check opportunities, the Opportunity Score ranking, at least three solutions per priority and the experiment thresholds.
What the skill makes the agent do
An Opportunity Solution Tree (OST) is a four-level map. At the top is one desired outcome, a single measurable metric. Beneath it sit customer opportunities, which are problems rather than features. Below those are possible solutions, and at the bottom are experiments that test whether a solution addresses the opportunity.
When the skill is active, the agent works through the method in order. It confirms or helps word a single outcome, identifies 3 to 7 customer opportunities from the research you provide, ranks them with an Opportunity Score (Importance × (1 − Satisfaction), with both values normalised to 0–1), and generates at least three solutions for each top opportunity. For the most promising solutions it suggests 1–2 fast experiments with a hypothesis, method, metric and success threshold. It then presents the whole tree hierarchically.
Install the skill from the ZIP
Download the ZIP for this skill from this page. The package contains one file, SKILL.md. Unzip it so that the skill folder sits in one of the locations below.
For a personal setup, unzip into ~/.claude/skills/. For a single project, unzip into the project's .claude/skills/ folder so it applies only there.
You can also upload the skill in the Claude apps under Settings → Capabilities → Skills.
- Personal: ~/.claude/skills/opportunity-solution-tree/
- Project only: .claude/skills/opportunity-solution-tree/ inside your project
~/.claude/skills/opportunity-solution-tree/SKILL.mdTrigger the skill with a natural request
You do not need a special command. The skill's description says to use it when structuring discovery work, mapping opportunities to solutions, or deciding what to build next, so ask for those things in plain language. Give the agent the outcome you are pursuing and the research you already have.
The skill expects three kinds of input: a desired outcome or business metric to improve, customer research data such as interviews, surveys, analytics or feedback, and optionally existing opportunities or solution ideas. The more concrete your research, the more useful the tree.
- "Build an opportunity solution tree for increasing 7-day retention. Here are my interview notes: ..."
- "We have these support tickets and survey answers. What opportunities should we focus on before choosing what to build next?"
- "Map our existing solution ideas for onboarding under the outcome of activated new users, then suggest experiments for the top two."
- "Help me structure our discovery work for next quarter using an opportunity solution tree."
What a good result looks like
Check that the outcome is a single metric, that the opportunities are written from the customer's view (for example "I struggle to..." or "I wish I could..."), and that each prioritised opportunity has at least three solutions. Each experiment should name a hypothesis, method, metric and success threshold.
The skill's method says to save the result as markdown when it is substantial, so ask for a file if you want to keep the tree and update it weekly as you learn.
- Outcome: one measurable metric, not a list of goals
- Opportunities: 3 to 7, grouped, focused on the top 2 to 3
- Solutions: at least 3 per prioritised opportunity, not just the first idea
- Experiments: 1 to 2 per promising solution, each with a success threshold
What the skill needs and safety notes
The skill lists no requirements. It does not run scripts and does not use the network, according to its listing. Its only file is SKILL.md, which contains instructions and no executable code.
Because the skill is only instructions, it cannot check your research for accuracy. The agent builds the tree from what you give it, so the output is only as reliable as your inputs.
Tips for better trees
Start from one outcome. The skill's principles say to focus the tree on a single desired outcome, so split broader goals into separate trees.
Describe problems, not features. The skill asks for opportunities framed as customer problems. If you paste a feature request, ask the agent to restate it as the underlying need.
Include the trio's view. The method calls for ideas from the product manager, designer and engineer together, so tell the agent which perspectives to weigh.
What people use it for
- Structuring a quarter of discovery work around one retention or activation outcome.
- Sorting support tickets and interview notes into customer opportunities before choosing features.
- Mapping existing solution ideas under the right opportunity to see which have no evidence behind them.
- Planning fast experiments with clear success thresholds for a product trio.
- Teaching the Opportunity Solution Tree method to a team or student.
Frequently asked questions
Is the opportunity solution tree agent skill free?
The skill is published under the MIT licence, as stated on this page. Check the licence file in the source repository for full terms before reuse.
What do I need to install the opportunity-solution-tree skill?
The listing lists no requirements. Download the ZIP, unzip it into ~/.claude/skills/ or a project's .claude/skills/ folder, or upload it under Settings → Capabilities → Skills in the Claude apps.
Does the skill run scripts or use the internet?
According to the listing, it runs no scripts and uses no network. The package contains only SKILL.md, which is a set of instructions.
Do I need customer research before using it?
Yes. The framework depends on real opportunities from interviews, surveys, analytics or feedback. Without that data the output would be guesswork.
Can it write product specs or code?
No. The skill only structures discovery thinking. For a finished spec or code you need a different tool or skill.
How is this different from just asking an AI for ideas?
The skill enforces the Opportunity Solution Tree structure: one outcome, opportunities before solutions, at least three solutions per opportunity and experiments with thresholds. Its method is based on Teresa Torres' Continuous Discovery Habits.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you run continuous discovery or want to learn the Opportunity Solution Tree method, and you have real customer research to feed it. Pick something else if you have no research data yet, or if you need a finished product spec or working code.
Download the skill