# Six Thinking Hats > Activate when: a meeting keeps going in circles, people have different agendas, someone says 'we need structure' or 'let's look at this from all angles,' a proposal needs evaluation from multiple sides, or creative ideas keep getting killed before development. Do NOT activate when: the question has a clear empirical answer (use critical-thinking instead), or time is under 10 minutes and no setup is possible. More: deciqai.com/s/six-thinking-hats - Skill: `deciqai/six-thinking-hats` (Agent Skill, multi-file: 4 files) - Install (CLI): `npx skillmds@latest add deciqai/six-thinking-hats` - Raw SKILL.md: https://api.skillmd.com/api/skills/deciqai/six-thinking-hats/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Productivity - Author: deciqAI (https://skillmd.com/u/deciqai) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/deciqai/six-thinking-hats --- # Six Thinking Hats ## Overview **Six Thinking Hats** (Edward de Bono, 1985) prevents the most common meeting failure: different thinking modes colliding simultaneously. The fix is *parallel thinking* — everyone uses the same mode at the same time. Six hats: **White** (data/facts), **Red** (gut/emotions), **Black** (caution/risk), **Yellow** (optimism/value), **Green** (creativity/alternatives), **Blue** (process/meta). ABB CEO Percy Barnevik cut 12-hour cross-cultural management meetings to under 2 hours after adoption. IBM and DuPont embedded the method in leadership development. **Compose with neighbors:** [lateral-thinking](../lateral-thinking/SKILL.md) populates Green Hat. [critical-thinking](../critical-thinking/SKILL.md) runs inside Black Hat. [mece](../mece/SKILL.md) structures the White Hat data phase. ## When to Use **Use when:** group decision degenerates into debate; proposal needs multi-angle evaluation; creative ideas get killed by habitual criticism; someone says *"let's look at this from all angles," "we keep going in circles," "we need structure"*; individual high-stakes decision needs forced multi-angle review; a platform-shift bet needs balanced evaluation (e.g. *"should we go all-in on AI agents / AI-native," "is our AI capex justified," "how do we weigh AI adoption risk vs. AI-native competition"*). **When NOT:** deep specialist analysis required (use hats for framing only); group too political (hat roles used as cover); time under 10 minutes; question is purely factual (run [critical-thinking](../critical-thinking/SKILL.md)). ## Coaching Novices (Adaptive Front Door) - **Engine mode:** user has a specific decision ready → run The Process directly. - **Coach mode:** user is unfamiliar or has no concrete case → guide step by step. In Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop. 1. **One-line what-it-is.** Six Thinking Hats separates discussion into six modes so everyone explores each angle *together* instead of arguing from different angles simultaneously. 2. **Check fit** — if the group is too political, question is purely factual, or time is under 10 minutes, redirect or modify. 3. **Elicit the specific decision or proposal.** "We need better meetings" is not workable; "should we launch feature X in Q3 or delay to Q4" is. > **[WAIT — do not advance until user responds]** 4. **One hat at a time.** Walk through each hat in sequence, posing the hat's key question and waiting for input before moving to the next. > **[WAIT — do not advance until user responds]** 5. **Close by naming which hat revealed the most unexpected insight** — the angle the team would have missed in an unstructured discussion. > **[WAIT — do not advance until user responds]** ## The Process Six phases → **Six Hats Analysis**. **Stop rule:** if Blue Hat close reveals no dominant direction, identify the most contested hat phase and schedule a focused session on it alone rather than forcing a decision. 1. **Blue Hat opens (5 min):** Define the question, hat sequence, time limits, and rules (no switching modes during a hat phase). 2. **White Hat — facts and data (10 min):** State known facts only. Distinguish confirmed data from soft data. List information gaps explicitly. 3. **Yellow Hat — optimism and value (8 min):** Generate specific benefits with logical basis. Run before Black to prevent habitual pessimism dominating. 4. **Black Hat — caution and risk (10 min):** Identify specific, evidence-based risks. "I don't like it" is Red Hat. "The API will fail under rate limits because X" is Black Hat. Do not abbreviate. 5. **Green Hat — creativity and alternatives (8 min):** Generate alternatives — no evaluation. Facilitator must actively block Black Hat intrusions. 6. **Red Hat — gut feeling (2 min, 30 sec per person):** One sentence per person, no justification required or permitted. 7. **Blue Hat closes (5 min):** Synthesize decision, action items, most critical Black Hat risk, most promising Green Hat direction. ### Output: Six Hats Analysis ```markdown # Six Hats Analysis: ## Setup (Blue Hat): Question | Hat sequence ## White Hat: Confirmed facts | Information gaps ## Yellow Hat: ## Black Hat: | Most critical: <...> ## Green Hat: ## Red Hat: : ## Blue Hat Close: Decision | Actions | Critical Black risk | Best Green direction ``` *→ Method in Action: [ABB Global Management Meetings (1990s)](examples/abb-global-management-meetings-1990s.md)* *→ 2026 lens: [Betting the Roadmap on AI Agents (2024–2026)](examples/betting-the-roadmap-on-ai-agents-2024-2026.md)* ## Hat Packs | Domain | White | Black | Yellow | Green | Note | |---|---|---|---|---|---| | Investment | metrics/comparables | thesis risks/exit obstacles | base-case return | alternative deal structures | Red Hat critical — legitimizes gut reaction | | Product feature | usage data/research | tech debt/adoption risk | user value/revenue | MVP alternatives | — | | Crisis | facts only | risks first | defer Yellow | alternatives | Red Hat runs first; Blue sequence = W→B→G→Blue | ## Applying It Well - **Blue Hat opening is the most critical 5 minutes.** Without explicit sequence, time limits, and rules, debate resumes by hat 3. - **Black Hat is the most valuable and most compressed hat.** Resist pressure to cut it "to stay positive." - **Run Yellow before Black** — establishes value worth protecting before naming threats. - **Green Hat needs active protection.** One evaluative comment collapses the whole phase. - **Red Hat is information, not complaint.** One sentence per person, no justification, no rebuttal. - **Blue Hat close is the deliverable.** Always allocate 5 minutes regardless of time pressure. *→ Primary sources: [references/sources.md](references/sources.md)* ## Common Rationalizations **[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.** | Fake move | Reality | |---|---| | [D] **Compressing Black Hat "to keep things positive."** | Black Hat is the session's risk register. Compressing it means deciding without the most important input. | | [D] **Using Red Hat as an argument channel.** | Red Hat is 30 seconds, one sentence, no justification. Extended statements are Black/Yellow in disguise. | | [D] **Fixed universal hat sequence regardless of context.** | Crisis → White-Black-Green-Blue. Strong internal advocates → Black before Yellow. Sequence must match decision type. | | [D] **No Blue Hat close — calling the session "done."** | Without Blue Hat close, the session produced discussion, not analysis. | | [D] **Facilitator expressing content preferences during a hat phase.** | Facilitator power is structural only. Content signals collapse the parallel thinking contract. | | [D] **Evaluating Green Hat ideas during the Green Hat phase.** | Green Hat ideas are candidates, not proposals. One evaluative comment collapses the whole phase. | | [D] **Only running hats natural to the team's culture.** | The method's value is forcing under-used modes. Risk-averse teams over-invest Black; innovator cultures skip it. | | [D] **White Hat shading into interpretation.** | "Sales down 15% — clearly the product isn't working" is not White Hat. Interpretation belongs in Yellow/Black/Green. | | *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* | ## Red Flags - Facilitator expressed content preferences during a hat phase - Black Hat compressed or skipped "to keep energy positive" - Green Hat ideas evaluated during the Green Hat phase - No Blue Hat close — session ended without a synthesized conclusion - Red Hat became an extended argument, not a one-sentence gut response - Same sequence used regardless of decision type ## Verification - [ ] Blue Hat opening: question defined, sequence set, time limits and rules stated - [ ] White Hat: confirmed vs. soft data distinguished; information gaps listed - [ ] Yellow Hat: logically grounded value statements (not wishful) - [ ] Black Hat: specific evidence-based risks (not emotional objections); not abbreviated - [ ] Green Hat: zero evaluation during the phase; alternatives captured unfiltered - [ ] Red Hat: one sentence per person, no justification, no rebuttal - [ ] Blue Hat close: decision or next action named; critical Black risk and best Green direction identified --- *Part of **deciqAI Knowledge Skills** — 237 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. **See it run → https://www.deciqai.com/s/six-thinking-hats** · Built by deciqAI · github.com/deciqAI · Contributions welcome.* *Agents: latest version & machine-readable metadata → https://www.deciqai.com/s/six-thinking-hats.json*