# How the Strategy-Red-Team Skill Stress-Tests Product Plans > Learn how the strategy-red-team skill stress tests product plans by challenging assumptions, analyzing claims, and ranking failure modes. Improve your product strategy. - Repository: [Pawel Huryn/pm-skills](https://github.com/phuryn/pm-skills) - Tags: deep-dive - Published: 2026-07-01 --- **The strategy-red-team skill rigorously challenges product requirements by extracting load-bearing assumptions, steel-manning then attacking each claim, and ranking failure modes by impact, likelihood, and cheapness to test.** The strategy-red-team skill is a structured red-team exercise defined in the `phuryn/pm-skills` repository that transforms PRDs, roadmaps, and strategic plans into prioritized, test-oriented risk assessments. By systematically dissecting assertions and surfacing the cheapest high-impact validation methods, it ensures teams validate critical assumptions before committing substantial resources. ## Extracting Load-Bearing Assumptions According to [`pm-execution/skills/strategy-red-team/SKILL.md`](https://github.com/phuryn/pm-skills/blob/main/pm-execution/skills/strategy-red-team/SKILL.md) (lines 20-22), the skill begins by parsing the input document to inventory every asserted claim about users, market size, constraints, mechanisms, or timelines. It applies a strict filter to retain only **load-bearing assumptions**—claims whose falsity would cause the entire plan to collapse. Minor uncertainties are discarded so the analysis focuses exclusively on existential risks. ## Steel-Manning Before Attacking For each retained claim, the skill applies a steel-man methodology. As specified in [`SKILL.md`](https://github.com/phuryn/pm-skills/blob/main/SKILL.md) (lines 22-24), it first articulates the strongest possible justification for the claim, then attacks that strengthened version. This approach prevents straw-man arguments and ensures the product plan survives only if it withstands the most charitable interpretation of its underlying logic. ## Defining Concrete Failure Modes The skill translates each load-bearing claim into a precise, falsifiable condition using the "Fails if …" format. This transformation, detailed in [`SKILL.md`](https://github.com/phuryn/pm-skills/blob/main/SKILL.md) (lines 24-26), replaces vague concerns with testable statements that explicitly define the threshold of failure. ## Ranking Risks by Impact, Likelihood, and Cheapness Each failure mode receives a composite score based on three dimensions defined in [`SKILL.md`](https://github.com/phuryn/pm-skills/blob/main/SKILL.md) (lines 26-28): - **Impact** — The severity of consequences if the claim proves false. - **Likelihood** — The probability that the claim is incorrect. - **Cheapness to test** — The effort required to obtain disconfirming evidence this week. The product of these three factors surfaces the **cheapest high-impact test** as the top priority, ensuring teams address existential risks with minimal validation cost. ## Specifying Evidence and Kill Criteria For every ranked "kill-assumption," the skill prescribes specific actionable items per [`SKILL.md`](https://github.com/phuryn/pm-skills/blob/main/SKILL.md) (lines 30-34): - **Evidence to get this week**: The exact data query, user interview, or experiment required. - **Kill criterion**: The quantitative threshold at which the plan must be halted or revised. - **Cheapest test**: The smallest experiment capable of validating or invalidating the assumption within days. ## Using the /red-team-prd Command The skill integrates into product workflows through the `/red-team-prd` command defined in [`pm-execution/commands/red-team-prd.md`](https://github.com/phuryn/pm-skills/blob/main/pm-execution/commands/red-team-prd.md) (lines 24-34). This command accepts a plan description, automatically feeds it into the strategy-red-team skill, and returns a formatted analysis. Example invocation: ```text /red-team-prd Prioritize AI onboarding — activation is our bottleneck ``` According to [`red-team-prd.md`](https://github.com/phuryn/pm-skills/blob/main/red-team-prd.md) (lines 54-66), the command returns a markdown block organized for immediate consumption and offers follow-up actions such as converting top risks into experiments or running complementary pre-mortems. ## Output Format and Structure The result is rendered in a screenshot-friendly markdown block specified in [`SKILL.md`](https://github.com/phuryn/pm-skills/blob/main/SKILL.md) (lines 38-56). The output separates: - **Top Kill-Assumptions**: Ranked by the impact-likelihood-cheapness score. - **What's Well-Reasoned**: Portions of the plan supported by strong evidence. - **What I Couldn't Assess**: Gaps where the plan lacked sufficient detail to evaluate. Example output: ```markdown ## Red-Team: Prioritize AI onboarding — activation is our bottleneck ### Top Kill-Assumptions (ranked) - **Claim:** Activation will increase by 30% after the new AI flow. - **Fails if:** Activation stays below the current 5% growth rate. - **Evidence to get this week:** Run a pilot with 50 users and measure activation lift. - **Kill criterion:** < 10% lift in the pilot. - **Cheapest test:** A/B test of the AI flow on the pilot cohort. ### What's Well-Reasoned The existing data shows a strong correlation between onboarding videos and activation, so the AI flow builds on a proven channel. ### What I Couldn't Assess No data on how the AI model scales under heavy load. ``` ## Summary - The strategy-red-team skill in `phuryn/pm-skills` converts product plans into testable risk assessments by isolating **load-bearing assumptions** from minor uncertainties. - It employs **steel-manning** to challenge the strongest version of each claim, avoiding straw-man critiques. - Risks are ranked by multiplying **impact**, **likelihood**, and **cheapness to test**, prioritizing validations that are both high-stakes and low-cost. - Each top risk includes a **kill criterion** and the **cheapest test** to run this week, ensuring actionable next steps. - Invoke the analysis using the `/red-team-prd` command, which formats results for immediate sharing and follow-up experimentation. ## Frequently Asked Questions ### What makes an assumption "load-bearing" in the strategy-red-team skill? An assumption is load-bearing if its falsity would cause the entire plan to collapse. The skill filters out minor uncertainties to focus exclusively on existential risks that threaten the strategy's core validity, as defined in the extraction phase of [`SKILL.md`](https://github.com/phuryn/pm-skills/blob/main/SKILL.md). ### How does the strategy-red-team skill differ from a standard pre-mortem? While pre-mortems imagine a future where the project failed and work backward to identify causes, the strategy-red-team skill steel-mans current assumptions and attacks them directly. It produces ranked, falsifiable conditions with specific kill criteria and cheap tests, whereas pre-mortems typically generate a broader list of potential failure modes without the same mathematical prioritization rigor. ### What is the "steel-man" approach mentioned in the skill documentation? Steel-manning is the opposite of straw-manning: it involves articulating the strongest possible version of a claim before attempting to refute it. The skill attacks only this strengthened version to ensure the product plan can survive genuine scrutiny, not just weak interpretations of its logic, as implemented in [`SKILL.md`](https://github.com/phuryn/pm-skills/blob/main/SKILL.md) (lines 22-24). ### How is the ranking score calculated for failure modes? The ranking score is the product of three factors: **impact** (severity if wrong), **likelihood** (probability of being false), and **cheapness to test** (effort required to validate). This multiplication surfaces risks that are both dangerous and easy to verify, maximizing the value of immediate testing efforts according to the scoring logic in [`SKILL.md`](https://github.com/phuryn/pm-skills/blob/main/SKILL.md) (lines 26-28).