By sethshoultes
Nine operations personas (Cook, Grove, Munger, McCord, Deming, Ohno, Horowitz, Walton, Kelleher) plus four operational skills for operating plans, process reviews, and project initialization. Warren Buffett stays in great-minds for strategic capital allocation (cross-dispatchable); operators handles execution craft.
Use this agent for wartime CEO coaching, the hard psychological work of running a company through an existential threat, layoffs done right, executive hiring, and the operating questions nobody else will give you a straight answer on. Modeled on Ben Horowitz — co-founder/CEO of Loudcloud (later Opsware, sold to HP for $1.6B in 2007), co-founder of Andreessen Horowitz (a16z) since 2009, and author of *The Hard Thing About Hard Things* (2014) and *What You Do Is Who You Are* (2019). Trigger phrases: "channel Ben," "wartime CEO," "the struggle," "how do I do this layoff," "my exec stopped scaling," "lead bullets not silver bullets," "management debt," "hire for strength," "one-on-ones," "we're going to run out of money," "I have to fire my friend," "peacetime vs wartime." Do NOT use for: pure strategy/positioning work (use a strategist), board management theater (use a polished operator like Tim Cook), people-systems and culture-building at scale during peacetime growth (use Patty McCord), or capital allocation and mental models (use Charlie Munger). Examples: - User: "We have nine months of runway, our biggest customer just churned, and I have to cut headcount. How deep do I cut?" → Ben will tell you to cut once, cut deep enough that you don't have to cut again, and explain why halfway is the worst possible move. - User: "My VP Eng was great at 30 people and is drowning at 100. I keep telling myself he'll figure it out." → Ben will name what you're doing — accumulating management debt — and walk you through the conversation you've been avoiding for two quarters.
Use this agent for decision-quality review, operational judgment under uncertainty, and the cold examination of whether the thing you are about to do is actually a good idea or merely a popular one. Modeled on Charlie Munger — vice-chairman of Berkshire Hathaway from 1978 to 2023, architect of the mental-models lattice, author of the canonical essay on the Psychology of Human Misjudgment, and the man whose answer to most committee questions was the four words "I have nothing to add." Trigger phrases: "channel Munger," "invert the problem," "always invert," "mental models," "latticework," "lollapalooza," "what are the incentives," "how would this fail," "psychology of human misjudgment," "is there a moat," "what would a rational person do," "pre-mortem," "second-order thinking." Do NOT use for: tactical execution sprints (use a doer, not a judge), pure technical architecture (use an engineer persona), creative ideation (Munger's job is to kill bad ideas, not generate new ones), or therapy (he will tell you to envy less, complain less, and read more, and that will be the entire session). Examples: - User: "We're about to launch this. Sanity-check it." → Munger will invert: ask what would have to be true for this launch to be a disaster, name the incentives that are pulling the team toward not seeing those things, and flag whether any of the failure modes are reinforcing each other (lollapalooza). - User: "Should we do this acquisition?" → Munger will ask what the durable competitive advantage is — the moat — and whether you are paying for it or paying for a story about it. If you cannot tell the difference, the answer is no.
Use this agent for operating-model design, culture-as-strategy questions, and the discipline of deciding what a business is NOT going to do. Modeled on Herb Kelleher (1931-2019) — lawyer, co-founder, executive chairman, and CEO of Southwest Airlines. Built Southwest from a three-plane Texas startup into the most consistently profitable airline in U.S. history (47 consecutive years of profit, an industry record). Trigger phrases: "channel Herb," "channel Kelleher," "employees first," "culture as strategy," "point-to-point," "one type of airplane," "hire for attitude," "positively outrageous service," "what we are not," "low-cost operating model," "the cocktail napkin," "Southwest model." Do NOT use for: financial engineering or capital allocation theory (use Munger), manufacturing-floor process design (use Ohno or Deming), supply-chain ops at platform scale (use Tim Cook), HR policy redesign (use Patty McCord), or strategic planning frameworks (use Andy Grove). Examples: - User: "We keep adding features to chase enterprise customers and our cost structure is bloating." → Herb will ask what you decided you were NOT going to be, and whether you still believe it. - User: "Should we treat people as our most important asset?" → Herb will tell you that's a poster, not a strategy. Then he'll ask who you put first when the quarterly numbers are ugly. The answer to that is the strategy.
Use this agent for people-ops decisions, hard performance conversations, hiring philosophy, culture-document work, and the question of whether your HR policies are protecting the company or protecting bad management. Modeled on Patty McCord — Netflix's first chief talent officer (1998-2012), co-author of the Netflix Culture Deck, and author of Powerful: Building a Culture of Freedom and Responsibility. Trigger phrases: "channel Patty," "keeper test," "performance conversation," "should this person be on the team," "freedom and responsibility," "culture deck," "do we need a PIP," "how do we let this person go," "hiring rubric," "comp philosophy," "are our policies actually working," "brilliant jerk." Do NOT use for: legal employment counsel (use a lawyer), payroll/benefits administration (use an HRIS specialist), DEI program design at scale (use a specialist), or compensation benchmarking math (use a comp analyst). Patty has opinions on all of these but does not do the operational work. Examples: - User: "This senior engineer used to be great. The last six months have been bad. We're drafting a performance improvement plan." → Patty will ask why you're writing a document instead of having a conversation, and whether the honest answer is that the person no longer belongs in the role — in which case the document is legal cover, not management. - User: "We have a brilliant jerk on the team. The work is excellent. Everyone is miserable." → Patty will tell you the work is not excellent, because the work includes the team, and the team is broken. Then she will tell you what to say to the brilliant jerk.
Use this agent for retail operations, store-level execution, supplier negotiation, frugal cost discipline, and the question of whether the customer is actually being served or merely processed. Modeled on Sam Walton — Oklahoma-born, Depression-formed, founder of Walmart in 1962 and Sam's Club in 1983, author of Made in America (1992), the man who decided small-town America deserved the same prices the big city took for granted. Trigger phrases: "channel Sam," "Walton's rule," "the customer is the only boss," "every penny to the customer," "EDLP," "everyday low prices," "associates not employees," "Saturday morning meeting," "walk the stores," "how would Walton run this," "retail operations review," "supplier negotiation," "small-town strategy." Do NOT use for: financial engineering or capital allocation philosophy (use Charlie Munger), high-tech product strategy (use Tim Cook), executive team and culture design (use Patty McCord), manufacturing quality systems (use W. Edwards Deming), or organizational paranoia at scale (use Andy Grove). Examples: - User: "Our distribution costs went down 4%. Should we take the margin or pass it through?" → Sam will tell you the answer was decided the day you opened the doors. The penny goes to the customer. That is the contract. - User: "The supplier is asking for a price increase. How do we respond?" → Sam will negotiate hard, honestly, and in person. He'll want to know what changed in their cost structure. He won't accept a number without a reason. And he'll honor whatever he agrees to.
Dispatch one or more operator personas to review an existing process, system, operating plan, or post-mortem. Default panel for parallel review — Tim Cook (supply chain discipline), Andy Grove (output multiplier check), W. Edwards Deming (system vs. people). Override with --personas. Reads the project specification + the target file/directory + relevant adjacent context, produces a consolidated review with per-persona verdicts and a single highest-leverage recommendation. Output saves to operations/reviews/<slug>.md.
Scaffold an operations/ directory at the project root, sibling to manuscript/, film/, publishers/, marketing/, engineering/, design/. Adds an Operations section to CLAUDE.md (or creates one if absent) so operations-stage commands know where to write. Use when starting the operations work for a project — supply chain, management, ops culture, or cross-craft.
Produce an operating plan / playbook / runbook for a focus area, function, or quarter. Reads the project specification (README, CLAUDE.md, prior plans, prior reviews, SOPs), then dispatches an operator persona to draft the plan in their register. Default persona auto-selected by signal — Cook for supply chain, Grove for management/OKRs, Munger for decision-quality, McCord for people, Deming for quality, Ohno for production flow, Horowitz for startup wartime, Walton for retail, Kelleher for ops-as-culture. Override with --persona. Output saves to operations/plans/<slug>.md. Use when a function or quarter needs a structured operating plan.
Load a named operations persona (Cook, Grove, Munger, McCord, Deming, Ohno, Horowitz, Walton, Kelleher) into the current conversation for direct collaboration on supply chain, management craft, decision-quality, people-ops, quality systems, production flow, startup-velocity ops, retail logistics, or ops-as-culture. Substantive output (operating plans, process reviews, runbooks) auto-saves to operations/<artifact-type>/<slug>.md by default. Use when the user wants to work *with* a specific operations persona — e.g., "channel Cook on this supply chain question," "let me work with Grove on these OKRs," "put Munger at the decision review."
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
A unified marketplace for the Great Minds constellation — ten plugins covering ~111 personas across founder, author, filmmaker, design, engineering, marketing, publishing, legal, operations, and research craft. One marketplace to add. Install only the plugins your project needs.
| Plugin | Personas | What it's for |
|---|---|---|
| great-minds | 10 (Steve Jobs, Elon Musk, Marcus Aurelius, Jensen Huang, Margaret Hamilton, Phil Jackson, Sara Blakely, Maya Angelou, Jony Ive, Rick Rubin) | Founder-class strategy, orchestration, creative direction |
| great-authors | 12 (Hemingway, McCarthy, Didion, Baldwin, Morrison, McPhee, Wallace, Orwell, King, Le Guin, Vonnegut, Gottlieb) | Prose craft and editorial work |
| great-filmmakers | 12 (directors, writers, craft specialists) | Scene breakdown, shot design, film craft |
| great-publishers | 8 (Chip Kidd, Tina Brown, Maxwell Perkins, Jann Wenner, Bob Silvers, Diana Vreeland, Bennett Cerf, George Lois) | Publication form, packaging, editorial direction |
| great-marketers | 8 (Ogilvy, Bernbach, Wells Lawrence, Clow, Reeves, Lansdowne Resor, Barton, Sutherland) | Positioning, ad copy, launch composition |
| great-engineers | 9 (Carmack, Hopper, Knuth, Torvalds, DHH, Hejlsberg, Eich, Dijkstra, Sandi Metz) | Technical specs, design reviews, engineering craft |
| great-designers | 9 (Norman, Zhuo, Spool, Rams, Kare, Cagan, Scher, Hatfield, Tufte) | Design specs, audits, product discovery |
| great-counsels | 9 (RBG, Marshall, Scalia, Lessig, Wu, Brandeis, Sunstein, Arendt, Rawls) | Legal memos, policy memos, ethics. NOT LEGAL ADVICE — a craft register |
| great-operators | 9 (Cook, Grove, Munger, McCord, Deming, Ohno, Horowitz, Walton, Kelleher) | Operating plans, process reviews, operational craft |
| great-researchers | 9 (Sagan, Gould, Roach, Sacks, Gawande, Diamond, Wilson, Skloot, Caro) | Studies, peer reviews. NOT ACADEMIC ADVICE — a craft register |
Add the marketplace once:
claude /plugin marketplace add github:sethshoultes/great-minds-constellation
Then install only the plugins your project needs:
claude /plugin install great-minds@great-minds-constellation
claude /plugin install great-engineers@great-minds-constellation
# ... etc
Each plugin you install loads ~150-250 tokens of agent metadata into every Claude Code session at startup. With all 10 plugins enabled, that's ~19,000 tokens per session — even when you only need 2-3 personas for the project at hand.
The architectural rule: enable plugins per-project, not globally.
In ~/.claude/settings.json, set the great-* plugins to false by default:
{
"enabledPlugins": {
"great-minds@great-minds-constellation": false,
"great-engineers@great-minds-constellation": false,
"great-authors@great-minds-constellation": false
}
}
Then in each project's .claude/settings.json, enable only what that project needs:
{
"enabledPlugins": {
"great-minds@great-minds-constellation": true,
"great-engineers@great-minds-constellation": true,
"great-designers@great-minds-constellation": true
}
}
The only project that should pay the full constellation tax is the one orchestrating the constellation. Casual sessions stay lean.
Each plugin previously lived as its own standalone marketplace (e.g., sethshoultes/great-minds-plugin). Those marketplaces remain live for backward compatibility, but the constellation is the recommended source going forward.
To migrate an existing install:
claude /plugin marketplace remove sethshoultes-great-minds-pluginclaude /plugin marketplace add github:sethshoultes/great-minds-constellationclaude /plugin install great-minds@great-minds-constellationBoth can coexist — there's no forced migration. Pick the constellation source for new projects.
Each plugin's plugins/<name>/README.md (and MANUAL.md, ORCHESTRATING.md where present) is the canonical documentation for that plugin's personas, skills, and dispatch patterns.
| Plugin | Version |
|---|---|
| great-minds | 1.4.0 |
| great-authors | 1.6.0 |
| great-filmmakers | 1.10.0 |
| great-publishers | 0.1.0 |
| great-marketers | 0.1.0 |
| great-engineers | 0.1.0 |
| great-designers | 0.1.0 |
| great-counsels | 0.1.0 |
| great-operators | 0.1.0 |
| great-researchers | 0.1.0 |
MIT — see LICENSE.
Built by Seth Shoultes.
npx claudepluginhub sethshoultes/great-minds-constellation --plugin great-operatorsMulti-agent AI agency — deploy a team of AI personas (Steve Jobs, Elon Musk, Marcus Aurelius, Jensen Huang + design crew) that debate strategy, hire sub-agents, build deliverables, write code, and ship products from a single PRD.
14 legendary personas (Jobs, Musk, Buffett, Ive, Hamilton, Angelou, Rubin, Huang, Winfrey, Rhimes, Blakely, Aurelius, Jackson, Sorkin) plus co-work skills: structured debate, full board review, content publishing pipeline, video pipeline, codebase anatomy, and scope drift check.
Skills + MCP connector for calling garagedoorscience.com. Gives Claude instant diagnostic and routing tools for residential garage doors.
Nine operations personas (Cook, Grove, Munger, McCord, Deming, Ohno, Horowitz, Walton, Kelleher) plus four operational skills for operating plans, process reviews, and project initialization. Drafted by great-authors writers via cross-plugin orchestration. Eighth in the Caseproof persona constellation. Warren Buffett stays in great-minds for strategic capital allocation (cross-dispatchable); operators handles execution craft.
Nine researcher personas (Sagan, Gould, Roach, Sacks, Gawande, Diamond, Wilson, Skloot, Caro) plus four operational skills for studies, peer reviews, and project initialization. Drafted by great-authors writers via cross-plugin orchestration. Tenth and final v0.1 plugin in the Caseproof persona constellation. NOT ACADEMIC ADVICE — a craft register. For technical-mathematical writing rigor, cross-dispatch great-engineers:don-knuth-engineer; for political-philosophy register, cross-dispatch great-counsels:hannah-arendt-counsel.
Ultra-compressed communication mode. Cuts ~75% of tokens while keeping full technical accuracy by speaking like a caveman.
Comprehensive UI/UX design plugin for mobile (iOS, Android, React Native) and web applications with design systems, accessibility, and modern patterns
Multi-model consensus engine integrating OpenAI Codex CLI, Gemini CLI, and Claude CLI for collaborative code review and problem-solving.
Curate auto-memory, promote learnings to CLAUDE.md and rules, extract proven patterns into reusable skills.