geoff-woods-ai-thought-partner

name

geoff-woods-ai-thought-partner

description

Geoff Woods' AI-as-strategic-thought-partner system — the Three Skills (target the 20%, communicate via CRIT, remain the thought leader), the Interview Inversion (AI asks YOU questions), the Feedback Triad refinement loop, stakeholder simulation with reality calibration, AI boards of advisors, and the cognitive-sovereignty guard. Plus Tier O: the always-on CO-CREATION-CARD operator layer that makes the thought-partner posture this workspace's default.

Geoff Woods — AI Thought Partner OS

Geoff Woods is the founder of AI Leadership and author of the #1 bestseller The AI-Driven Leader. Before that he co-founded the training and consulting company behind The ONE Thing with Gary Keller and Jay Papasan, and was a C-level exec at a public company based in India (external bios identify it as Jindal Steel & Power, CGO, with a $750M → $12B market-cap arc — bio claims, not from the interview; see references/source-ledger.md). His system inverts the default AI relationship: "You are the thought leader. AI is your thought partner." Instead of asking AI questions like it's Google, he makes AI interview him — one question at a time — to pull context out of his head, elevate his thinking, and only then produce. Every output is "the bad answer" until the Feedback Triad has run. This skill encodes his full operating system, and its Tier O layer (CO-CREATION-CARD.md) installs the posture as this workspace's always-on default.

Available Workflows

Slash commands: /gw-crit /gw-thought-partner /gw-20-percent /gw-feedback-loop /gw-ai-board /gw-persona-flip /gw-simulate-room /gw-skill-to-agent /gw-cognitive-guard /gw-jam-fusion /gw-forge-brief /gw-voice-persona /gw-thinking-leverage /gw-content-pretest /gw-challenger /gw-custom-instructions /buyer-council (wrappers in .agent/workflows/, numbered to match the tables below).

Tier 1 — Foundation (Core Methodology)

Workflow	Command Intent	Produces
workflows/01-crit.md	Build a CRIT prompt for any problem	Deployment-ready Context/Role/Interview/Task prompt with the "what else" depth loop and delta-seeking task wording
workflows/02-thought-partner.md	Run a full strategic thought-partner session	Biggest-problem targeting → live interview inversion → non-obvious strategies → triad-iterated answer
workflows/03-twenty-percent.md	Prioritize the plate via 20/80	Starred 20% list, the ONE priority, and the AI-handoff column (what AI is ≥50% as good at)
workflows/04-feedback-loop.md	Elevate any first-pass output	"Bad answer" framing → structured Feedback Triad rounds → plateau detection → Challenger handoff

Tier 2 — Practitioner (Granular Tools)

Workflow	Command Intent	Produces
workflows/05-ai-board.md	Build a custom AI board of advisors	Skill-gap-derived advisor personas with permission rules + future-self member, deployable as markdown
workflows/06-persona-flip.md	Adversarially harden a deliverable	Serial persona rotation: Challenger → ideal customer → aggressive board member, with anti-sycophancy architecture
workflows/07-simulate-room.md	Pre-test a deliverable against real stakeholders	Personality profiles → simulated room reactions → adjustments → mandatory reality-calibration loop
workflows/08-skill-to-agent.md	Convert a tacit skill/process to an agent spec	Dominoes-gated, interview-extracted markdown an agent can run without you
workflows/09-cognitive-guard.md	Audit and protect thinking strength	Cognitive-sovereignty audit ("can I still fly?") + thinking-time-chair protocol
workflows/15-challenger.md	Fast standalone red-team on anything	The Challenger split from persona-flip: cracks/biases/assumptions + triad-on-the-critique, no plateau or audience needed
workflows/16-custom-instructions.md	Build a portable global AI operating profile	Interview-built master profile → paste-ready blocks per tool (ChatGPT/Claude/Copilot); workspace already carries this via Tier O

Tier 3 — Stacking (Cross-Expert Chains)

Workflow	Command Intent	Produces
workflows/10-jam-fusion.md	Co-create taste-bearing work	Woods interview front-end → /jam two takes side-by-side → gut verdict → triad on the winner
workflows/11-forge-brief.md	Turn raw intent into a forge-ready build brief	Dominoes-gated CRIT interview → markdown build brief + forge-os lane recommendation
workflows/12-voice-persona.md	Build a writing persona from samples	His exact recipe (20-50 samples → interview → markdown voice file), reconciled with voice-os for client voices
workflows/14-content-pretest.md	Pre-test a draft against a deep-ICP profile	simulate-room engine with an ICP loaded as the reader → per-section reactions, the exact drop-point line, identity triggers → calibration loop vs. real engagement
workflows/17-buyer-council.md	Convene a simulated buyer panel against any artifact (triage + council modes)	Front door /buyer-council: standing-roster reuse (councils/buyers/) → per-element KILL/REVISE/SHIP verdicts, dissent preserved, mandatory NO-FIT + economic-gatekeeper seats, Predicted Response Map, calibration.jsonl reality loop

Tier 4 — Leadership (The AI-Driven Leader layer)

Workflow	Command Intent	Produces
workflows/13-thinking-leverage.md	Coach a person (report/client/self) to own 100% of their job	"Show me your plan" one-on-one: they rank their own 20%, you help them see what they don't see; superpowers→20% match; networks-not-hierarchy org frame

Tier O — Operator Layer (always-on)

references/CO-CREATION-CARD.md — the fusion protocol card wired into CLAUDE.md. Two labeled lineages: Woods moves (interview-before-produce, bad-answer framing, Feedback Triad, 20% check, Challenger flip, context hunger) + Claude-native moves (memory-before-mouth, intent-score scaling, first-take-at-standard, parallel takes, push-back-don't-comply, options-with-recommendation, conduct-the-fleet). Dial: PARTNER / EXECUTE / OFF.

Stacking Guide

Pair with	When
/jam	Taste-bearing co-creation (native chain: /gw-jam-fusion)
/forge (forge-os)	The interview surfaces a repeatable tool worth building (/gw-forge-brief)
/voice-os	Writing personas for Farrice (voice-os is canonical) vs. new/client voices (/gw-voice-persona)
/dan-koe-ai-leverage	Creator-leverage angle on the same posture — Koe for content thinking, Woods for strategic decisions
/mark-kashef-ai-councils	The AI board needs structured multi-agent debate instead of persona simulation
/nate-b-jones-intent-engineering	The interview output should become a durable intent doc for agents
/cos	20% targeting feeds the daily brief / weekly board
/wargame-brief · /convene	The simulated room should escalate to a full council or wargame
icp-deep-canvasser	Pre-test content against the real reader: load a deep-ICP profile into simulate-room as the stakeholder (/gw-content-pretest)
/haynes-handshake-geoff-stakeholder	Buyer Council's cold-offer specialization — objection-inventory intake + COMPOSE/LAYER revision loop for a Haynes offer stack specifically

Quick Reference

• The Three Skills: (1) target a 20% use case ("if you aim it at something that doesn't matter, it doesn't matter"), (2) communicate via CRIT, (3) remain the thought leader (skeptical, iterating, never submitting slop).
• CRIT: Context (verbose, unedited, voice-dump; "when you think you've given enough, assume you have not — ask 'what else?'") → Role (vivid specific expert, never a category label) → Interview (the most important part: "ask me one question at a time, up to five, to gain deeper context") → Task (request the DELTA: non-obvious strategies, what-I-don't-know).
• The Feedback Triad: "Here's what I like / what I don't like / the top changes." Run it on every output; a prompt becomes a conversation.
• "This is the bad answer": the pre-read ritual. Every output is a first draft regardless of how good it sounds.
• Persona flip: at plateau ("best I can do ≠ best that can be done") rotate: Challenger → ideal customer → aggressive growth-minded board member.
• Dominoes: agentic AI is the 18th domino. Three Skills daily × 30 days first ("you haven't seen color yet"). Agents only for documented, repetitive, 20%-tied tasks — one at a time, made excellent.
• Job = skills + processes: anything undocumented in your head → CRIT interview → markdown → agent replicates it.
• Cognitive sovereignty: AI that answers your questions replaces your thinking; AI that questions your answers strengthens it. Self-test: "Can I still solve problems without it?" Pen-and-paper thinking time upstream of AI, always.
• AI-repricing rule: if AI is 50% as good as you at a task today, start handing it over — in 12 months it runs away with it. Durable human skills: strategic thinking, problem solving, communication, collaboration (people + machines), creativity.
• Conductor creativity: "You're no longer the player of one instrument. You're the conductor" — creativity relocates to vision-casting, role-casting, task design.
• Claims quarantine: his case-study outcomes ($300M save, board turnaround, $750M→$12B) are HIS stated figures — cite as his claims, never as verified facts.

Execution Prompts (structure-pure v2)

16 deterministic practitioner prompts — each carries an Output Contract, Output Skeleton, and Quality Gate. When a deliverable matches one, Read it and honor its contract instead of improvising the output shape.

• Geoff Woods — AI Board of Advisors — skills/geoff-woods-ai-thought-partner/references/prompts-v2/ai-board.md
• Geoff Woods — The Challenger (Standalone Red-Team) — skills/geoff-woods-ai-thought-partner/references/prompts-v2/challenger.md
• Geoff Woods — Cognitive Sovereignty Guard — skills/geoff-woods-ai-thought-partner/references/prompts-v2/cognitive-guard.md
• Geoff Woods — Content Pre-Test (ICP as the Room) — skills/geoff-woods-ai-thought-partner/references/prompts-v2/content-pretest.md
• Geoff Woods — CRIT Prompt Builder — skills/geoff-woods-ai-thought-partner/references/prompts-v2/crit.md
• Geoff Woods — Global Custom-Instructions Profile — skills/geoff-woods-ai-thought-partner/references/prompts-v2/custom-instructions.md
• Geoff Woods — Feedback Loop & Slop Kill — skills/geoff-woods-ai-thought-partner/references/prompts-v2/feedback-loop.md
• Build: [name] — skills/geoff-woods-ai-thought-partner/references/prompts-v2/forge-brief.md
• Geoff Woods — Jam Fusion Co-Creation — skills/geoff-woods-ai-thought-partner/references/prompts-v2/jam-fusion.md
• Geoff Woods — Persona Flip Adversarial Hardening — skills/geoff-woods-ai-thought-partner/references/prompts-v2/persona-flip.md
• Geoff Woods — Simulate the Room — skills/geoff-woods-ai-thought-partner/references/prompts-v2/simulate-room.md
• [Agent Name] — skills/geoff-woods-ai-thought-partner/references/prompts-v2/skill-to-agent.md
• Geoff Woods — Thinking-Leverage Leadership Protocol — skills/geoff-woods-ai-thought-partner/references/prompts-v2/thinking-leverage.md
• Geoff Woods — Strategic Thought-Partner Session — skills/geoff-woods-ai-thought-partner/references/prompts-v2/thought-partner.md
• Geoff Woods — 20% Targeting Ritual — skills/geoff-woods-ai-thought-partner/references/prompts-v2/twenty-percent.md
• Writing Persona: [subject] — skills/geoff-woods-ai-thought-partner/references/prompts-v2/voice-persona.md