outcome-hypothesis

Create or assess evidence-grounded, falsifiable business outcome hypotheses with readiness scoring and validation rules.

1.4k|284|Updated Nov 2, 2025
One-click install
npx skills add https://github.com/microsoft/hve-core --skill outcome-hypothesis
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: outcome-hypothesis
Source: https://github.com/microsoft/hve-core/tree/main/.github/skills/project-planning/outcome-hypothesis
Command: npx skills add https://github.com/microsoft/hve-core --skill outcome-hypothesis

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Teams often frame initiatives around deliverables like MVPs or proofs of concept instead of measurable business results, making it impossible to judge whether an investment is justified. This Skill converts technical ideas into falsifiable outcome hypotheses with numeric targets, baselines, and indicators, or rigorously assesses an existing hypothesis for investability.

Core Features & Use Cases

  • Create Mode: Drafts a full or provisional outcome hypothesis from gathered evidence, using a D1-D7 readiness scorecard to decide between Ready, Provisional, or Investigate paths.
  • Assess Mode: Evaluates a supplied hypothesis read-only against validation rules OH.0-OH.13, reporting a findings table, investability verdict, and confidence level without altering the source.
  • Governance & Handoff: Enforces privacy gates for individual-level measurement, a human-owned status lifecycle (Draft, Provisional, Committed), and an optional BRD handoff payload to the requirements-author skill.
  • Use Case: A product manager wants to justify a new onboarding feature. The skill scores available evidence, drafts a hypothesis stating the beneficiary, lagging KPI with baseline and target, timeframe, and measurement plan, then flags gaps that block investability.

Quick Start

Ask the agent to create an outcome hypothesis for your initiative, providing any discovery evidence such as meeting notes, analytics, or prior artifacts.

Frequently Asked Questions about outcome-hypothesis

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write a falsifiable outcome hypothesis for a product initiative?

Provide your discovery evidence and request create mode. The skill scores seven readiness pillars, then drafts a hypothesis with a beneficiary, numeric leading and lagging indicator targets, a timeframe, baselines, and falsification criteria, marking unsupported sections as explicit gaps.

How do I assess whether an existing outcome hypothesis is investable?

Supply the existing hypothesis document and request assess mode. The skill evaluates it read-only against rules OH.0 through OH.13, reports a findings table, and labels it not investable if rules on targets, timeframes, baselines, or the outcome chain fail.

What is the difference between create mode and assess mode?

Create mode authors a new hypothesis from gathered evidence and may persist it after confirmation. Assess mode preserves a supplied hypothesis unchanged, reports validation findings for every rule, and only offers a revised draft as a separate follow-up request.

When does the outcome hypothesis workflow stop for privacy review?

It stops when an individual-level measure uses personal or sensitive data. Scoring, drafting, and validation pause until a completed Privacy Planner result exists, or a documented review by a qualified privacy professional when the planner is unavailable.

When should I use requirements-author instead of an outcome hypothesis?

Use requirements-author when the outcome is already understood and the task is creating or governing a BRD or PRD. The outcome hypothesis skill handles earlier framing of measurable results and can hand off a committed artifact to requirements-author.