How do I install the Unknown skill?

Run npx skillmds add october-academy/unknown in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Unknown safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Unknown?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Unknown free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Unknown?

october-academy (@october-academy) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Unknown by october-academy — Agent Skill

Name: Unknown
Rating: 4.5 (4 reviews)
Author: october-academy

Unknown

This skill should be used when the user provides a strategy, plan, or decision document and wants to surface hidden assumptions and blind spots using the Known/Unknown 4-quadrant framework. Trigger on "known unknown", "4분면 분석", "blind spots", "what am I missing", "뭘 놓치고 있지", "전략 점검", "가정 점검", "quadrant analysis". For requirement clarification use interview; for content-vs-form reframing use metamedium.

october-academy fd980c7 3 files · 4.4 KB Updated Sep 9, 2026

File contents

Unknown: Strategy Blind Spot Analysis

Surface hidden assumptions and strategic blind spots using the Known/Unknown 4-quadrant lens.

When to Use

Strategy documents that look coherent but are hard to execute
Plans with unclear risk boundaries
Decision discussions where the team asks: "What are we missing?"

For requirement clarification, use interview. For content/form leverage decisions, use metamedium.

Core Lens

Classify insights into four quadrants:

Known Knowns (KK): validated facts and working mechanisms
Known Unknowns (KU): explicit uncertainties requiring experiments
Unknown Knowns (UK): existing but underused assets/capabilities
Unknown Unknowns (UU): unobserved risks requiring early-warning antennas

Default resource split (adjustable):

KK 60%: systematize
KU 25%: experiment
UK 10%: leverage
UU 5%: install antennas

Non-Negotiable Interaction Rule

Use AskUserQuestion for each round (R1/R2/R3). Ask with hypothesis options, not open free-text prompts. On Codex the equivalent tool is request_user_input; with no structured-question tool, present the same options as a numbered plain-text list and wait for the reply.

3-Round Pattern

R1 (3-4 questions): broad scan across all quadrants
R2 (2-3 questions): drill into weakest link discovered in R1
R3 (2-3 questions, optional): execution detail for top priorities

Cap total questions at 7-10.

Protocol

Phase 1: Intake

If a doc is provided, extract goals, assumptions, dependencies, and unstated constraints.

Phase 2: Context Sweep

Scan project context (README, plans, prior notes) to identify likely UK candidates (assets the team has but is not using).

Phase 3: Draft + R1

Create a provisional 4-quadrant draft, then run R1 AskUserQuestion (single batched call).

Phase 4: R2 Deepening

Design R2 only from R1 responses:

Compound bottlenecks
Surprising answers
"Other" selections

Phase 5: R3 Execution (optional)

If needed, confirm owner, timeline, stop-criteria, and promotion conditions.

Phase 6: Playbook Output

Produce a structured playbook using references/playbook-template.md.

Rules

Hypotheses as options
R2 derives from R1, R3 derives from R2
Include both what to start and what to stop
Every KU gets promotion condition and kill condition
Prefer small experiments over large speculative plans

Additional Resources

references/question-design.md
references/playbook-template.md

---name: unknowndescription: This skill should be used when the user provides a strategy, plan, or decision document and wants to surface hidden assumptions and blind spots using the Known/Unknown 4-quadrant framework. Trigger on "known unknown", "4분면 분석", "blind spots", "what am I missing", "뭘 놓치고 있지", "전략 점검", "가정 점검", "quadrant analysis". For requirement clarification use interview; for content-vs-form reframing use metamedium.--- # Unknown: Strategy Blind Spot Analysis Surface hidden assumptions and strategic blind spots using the Known/Unknown 4-quadrant lens. ## When to Use - Strategy documents that look coherent but are hard to execute - Plans with unclear risk boundaries - Decision discussions where the team asks: "What are we missing?" For requirement clarification, use **interview**. For content/form leverage decisions, use **metamedium**. ## Core Lens Classify insights into four quadrants: - **Known Knowns (KK)**: validated facts and working mechanisms - **Known Unknowns (KU)**: explicit uncertainties requiring experiments - **Unknown Knowns (UK)**: existing but underused assets/capabilities - **Unknown Unknowns (UU)**: unobserved risks requiring early-warning antennas Default resource split (adjustable): - KK 60%: systematize - KU 25%: experiment - UK 10%: leverage - UU 5%: install antennas ## Non-Negotiable Interaction Rule Use **AskUserQuestion** for each round (R1/R2/R3). Ask with hypothesis options, not open free-text prompts. On Codex the equivalent tool is `request_user_input`; with no structured-question tool, present the same options as a numbered plain-text list and wait for the reply. ## 3-Round Pattern - **R1 (3-4 questions)**: broad scan across all quadrants - **R2 (2-3 questions)**: drill into weakest link discovered in R1 - **R3 (2-3 questions, optional)**: execution detail for top priorities Cap total questions at **7-10**. ## Protocol ### Phase 1: Intake If a doc is provided, extract goals, assumptions, dependencies, and unstated constraints. ### Phase 2: Context Sweep Scan project context (README, plans, prior notes) to identify likely UK candidates (assets the team has but is not using). ### Phase 3: Draft + R1 Create a provisional 4-quadrant draft, then run R1 AskUserQuestion (single batched call). ### Phase 4: R2 Deepening Design R2 only from R1 responses: - Compound bottlenecks - Surprising answers - "Other" selections ### Phase 5: R3 Execution (optional) If needed, confirm owner, timeline, stop-criteria, and promotion conditions. ### Phase 6: Playbook Output Produce a structured playbook using `references/playbook-template.md`. ## Rules 1. Hypotheses as options 2. R2 derives from R1, R3 derives from R2 3. Include both `what to start` and `what to stop` 4. Every KU gets promotion condition and kill condition 5. Prefer small experiments over large speculative plans ## Additional Resources - `references/question-design.md` - `references/playbook-template.md`

Frequently asked questions