How the Opportunity Solution Tree Skill in pm-skills Structures Continuous Discovery

The Opportunity Solution Tree skill in pm-skills codifies Teresa Torres’s discovery framework into a four-layer hierarchy—mapping a Desired Outcome to Opportunities, Solutions, and Experiments—to prevent premature solution commitments and drive validated product decisions.

The Opportunity Solution Tree (OST) skill is a standalone reference in the phuryn/pm-skills repository designed to help product teams structure continuous discovery work. Unlike traditional feature backlogs, this skill implements Torres’s framework to ensure teams prioritize customer problems over predefined solutions. It operates as an agent-invocable module rather than a CLI tool, enabling AI assistants to guide users through structured discovery workflows based on the definitions in pm-product-discovery/skills/opportunity-solution-tree/SKILL.md.

Architecture of the Opportunity Solution Tree Skill

The skill organizes discovery work into a hierarchical tree structure that forces explicit validation at every level. According to the source documentation, this architecture prevents teams from "solutioning" too early by separating problem identification from solution generation.

The Four-Layer Hierarchy

The tree consists of four distinct layers defined in the skill configuration:

• Desired Outcome – The root node containing a single, measurable metric (e.g., "increase 7‑day retention to 40 %"). This typically derives from OKRs or product strategy and anchors the entire tree【SKILL.md†line‑16】.

• Opportunities – Second-level nodes representing real customer problems discovered through research. The skill recommends prioritizing these using the Opportunity Score calculated as Importance × (1 − Satisfaction)【SKILL.md†line‑18】.

• Solutions – Third-level nodes containing multiple ideas per opportunity. The framework mandates that the Product Trio (PM + Designer + Engineer) collaboratively generate at least three solutions before committing to any single approach【SKILL.md†line‑20】.

• Experiments – Leaf nodes defining fast, cheap tests that validate solution assumptions across value, usability, viability, and feasibility dimensions. The skill emphasizes "skin‑in‑the‑game" experiments that force concrete validation【SKILL.md†line‑22】.

Core Operating Principles

The SKILL.md file (lines 24‑31) establishes five governing principles that distinguish this approach from traditional roadmapping:

1. One outcome at a time – Maintain strict focus on a single metric to prevent scope dilution.
2. Opportunities, not features – Treat customer problems as the primary unit of work rather than output features.
3. Generate ≥ 3 solutions – Force ideation breadth before convergence to avoid premature commitment.
4. Iterative validation – Kill solutions that fail experiments and branch new ideas rather than persisting with disproven approaches.
5. Continuous update – Refresh the tree weekly as new research insights emerge.

Using the Opportunity Solution Tree Skill in pm-skills

The OST skill does not expose a dedicated CLI command. Instead, agents (such as Claude) invoke it automatically when prompts require structured discovery work.

Input Requirements

To trigger the skill properly, provide three core inputs defined at lines 36‑40 in SKILL.md:

1. A clearly defined desired outcome (measurable metric)
2. Raw research data (customer interviews, support tickets, analytics)
3. Optional: Pre-existing opportunities or solution ideas to incorporate into the tree

Step-by-Step Workflow

When invoked, the skill executes a six-phase process outlined in lines 41‑55:

1. Outcome definition – Confirm or establish the single metric driving the initiative.
2. Opportunity extraction – Parse research data to identify customer problems.
3. Prioritization – Score and rank the top 2‑3 opportunities using the Opportunity Score formula.
4. Solution generation – Brainstorm three distinct solutions per prioritized opportunity.
5. Experiment design – Propose rapid, low-cost tests for the most promising solutions.
6. Hierarchy output – Return the complete tree as a markdown structure for team alignment.

Code Examples and Implementation

Prompt Template for Agent Invocation

Use this template to trigger the skill through an AI assistant:

**Prompt to the assistant (using the OST skill)**

Build an Opportunity Solution Tree for the outcome: {{OUTCOME}}.

Research data:
{{RESEARCH_SNIPPETS}}

Provide:
1. A list of 3-5 opportunities (customer-centric phrasing).
2. For each opportunity, 3 solution ideas.
3. For each solution, one experiment (hypothesis, method, metric, success threshold).
4. The full tree in markdown hierarchy.

When the assistant receives this template, it calls the opportunity-solution-tree skill and follows the step-by-step process defined in the skill file.

Generated Output Structure

The skill returns a markdown hierarchy that visualizes the discovery tree:

## Opportunity Solution Tree – Increase 7‑day Retention to 40 %

- **Outcome**: Increase 7‑day retention to 40 %
  - **Opportunity 1**: Users abandon onboarding after step 3
    - **Solution A**: Reduce onboarding steps from 5 → 3
      - **Experiment**: A/B test shortened flow; metric = completion rate
    - **Solution B**: Add progress indicator
      - **Experiment**: Show indicator; metric = time‑on‑step
    - **Solution C**: Provide inline help tooltip
      - **Experiment**: Tooltip click‑through; metric = help‑usage rate
  - **Opportunity 2**: Users forget to return after first week
    - **Solution A**: Push notification reminder
      - **Experiment**: Send reminder; metric = day‑7 return rate
    - **Solution B**: In‑app gamified streak badge
      - **Experiment**: Badge rollout; metric = repeat‑login count
    - **Solution C**: Email recap of benefits
      ...

Source File Locations

The following files define the skill’s behavior and usage:

• pm-product-discovery/skills/opportunity-solution-tree/SKILL.md – Core definition of the OST skill, including the four-layer hierarchy, Opportunity Score calculation, and principles【SKILL.md†line‑16‑31】.
• pm-product-discovery/README.md – Lists the OST skill among other discovery-related capabilities in the module.
• README.md (repo root) – Explains that OST is a standalone reference invoked by agents rather than a CLI tool.

Summary

• The Opportunity Solution Tree skill in pm-skills implements Teresa Torres’s framework as a four-layer hierarchy: Desired Outcome → Opportunities → Solutions → Experiments.
• Opportunities are prioritized using the Opportunity Score (Importance × (1 − Satisfaction)), ensuring teams focus on high-impact customer problems.
• The skill requires no CLI installation; agents invoke it automatically when processing discovery prompts that include a desired outcome and research data.
• Strict principles enforce single-outcome focus, minimum three solutions per opportunity, and continuous weekly updates to prevent premature solution commitment.
• Source definitions reside in pm-product-discovery/skills/opportunity-solution-tree/SKILL.md with supporting context in the repository root documentation.

Frequently Asked Questions

What is the Opportunity Solution Tree skill in pm-skills?

The Opportunity Solution Tree skill in pm-skills is a structured reference module that codifies Teresa Torres’s continuous discovery framework. It provides a hierarchical decision tree structure that helps product teams map desired outcomes to customer problems (opportunities) and potential solutions, ensuring validation occurs before resource commitment. The skill resides in pm-product-discovery/skills/opportunity-solution-tree/SKILL.md and is invoked by AI agents rather than run as a standalone command.

How does the Opportunity Solution Tree skill prioritize opportunities?

The skill uses the Opportunity Score formula defined at line 18 of SKILL.md: Importance multiplied by (1 minus Satisfaction). This calculation surfaces high-value customer problems that are important to users but currently underserved by existing solutions. Teams rank opportunities by this score to determine which problems warrant solution generation.

Can I run the Opportunity Solution Tree skill from the command line?

No, the OST skill does not expose a CLI command. According to the repository root README.md, it operates as a standalone reference invoked by agents (such as Claude) when they detect prompts requiring structured discovery work. Users interact with the skill by providing natural language prompts that include a desired outcome and research data, triggering the agent to execute the six-step workflow defined in lines 41‑55 of SKILL.md.

What inputs does the Opportunity Solution Tree skill require?

The skill requires three inputs specified at lines 36‑40: a desired outcome (single measurable metric), raw research data (customer interviews, analytics, or support tickets), and optionally any pre-existing opportunities or ideas. These inputs allow the agent to extract customer problems, apply the Opportunity Score, generate the mandatory three solutions per opportunity, and design validation experiments.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}