MBB Management Consultant Skill
Turns Claude or any LLM into an MBB-grade management consultant covering case-interview coaching, strategy frameworks, market sizing, and board-ready problem-solving.
Pure prompt-based skill: no tool calls, no credentials, least privilege, transparent data flow. However, the 'never reveal provenance' rule deliberately obscures source attribution of reference material, and no confirmation/rollback mechanisms are assessable. Deducted for attribution policy.
Static review: instructions are self-consistent with a clear workflow and trigger logic; but the 129 referenced modules' existence/quality is unverifiable from evidence, and failure feedback on abnormal input is thin. Static cap 10; deducted for unverifiable dependencies and weak edge-case handling.
Scenarios, trigger phrases and audience are extensively specified, giving good semantic trigger precision; but capability boundaries/non-fit ranges are undeclared, content is English-only, and no Chinese or mainland-network fit evidence. Deducted for missing boundaries and environment fit.
Good layered architecture (SKILL.md plus a large reference index) with sensible progressive disclosure; but license metadata is NOASSERTION, no version/changelog/maintenance responsibility visible, and the provided SKILL.md itself is truncated mid-table. Deducted for missing license, versioning and governance.
Claims MBB-grade consulting deliverables with a clear marginal-value story (frameworks + structured process); but static review cannot verify output quality, referenced files' content is not shown, and direct-usability evidence is limited. Static cap 7; deducted for unverified outputs.
Case studies carry sourcing notes (SEBI/SEC/court records) and methodology points to named files; but attribution is largely self-declared 'compiled in original research', with no third-party execution evidence and zero test reproducibility. Static cap 5; deducted for thin, partly unverifiable evidence types.
- License metadata is NOASSERTION; verify redistribution terms before commercial use.
- The skill instructs concealment of reference-material provenance, weakening attribution integrity; independently verify case-study sources before citing them.
- Existence and quality of the 129 referenced modules were not verifiable in this static review; confirm file completeness before relying on them.
- English-only content with no declared Chinese support; core value depends entirely on model reasoning with no external service dependency, so mainland-China reachability is unaffected.
- No tests or CI evidence that key paths were validated; output quality rests on model capability alone — manually review all quantified conclusions.
What does this skill do, and when should you use it?
An open-source Claude skill and plugin that ships 129 reference modules to turn a large language model into a senior strategy consultant with hypothesis-driven, structured thinking. It covers MECE issue trees, the Pyramid Principle, case-interview coaching, industry intelligence (healthcare, financial services, energy, TMT, consumer, and more), plus specialist methods like due diligence, post-merger integration, and GenAI enterprise strategy. It installs with one command in Claude Code and also works with Codex, Gemini, Copilot, and Ollama by attaching SKILL.md. The project is free and explicitly states it contains only publicly sourced knowledge — no confidential consulting-firm material.
SKILL.md defines a 7-step workflow: pattern-match the brief to five 2025-2026 demand archetypes, clarify the problem statement, build a MECE issue tree with a Day 1 Answer, then draw on 129 deep reference files in references/ for quantitative analysis (unit economics, Fermi estimation, industry benchmarks), Pyramid Principle synthesis, value quantification (definite and derived), an implementation plan, and commercial contracting. It calls no external scripts or APIs — it is purely prompt-and-reference driven. Outputs include action-titled slide structures, executive summaries, issue trees, market sizings, case scoring on a 5-dimension 1-4 rubric, and 100-day transformation plans.
- MBA students and case candidates: have the AI play interviewer through profitability, market entry, or M&A cases and get scored on structure, hypothesis, quant, and synthesis
- Working consultants: quickly draft issue trees for declining-margin diagnoses and condense 200 facts into 5 storyline messages for decks and memos
- Founders and operators: facing a margin slide or market-entry decision, get board-ready structured recommendations using 80/20 and Pyramid Principle thinking
- M&A practitioners: pull on-demand modules for due diligence checklists, synergy modeling, SPA term benchmarking, and post-merger integration frameworks
- Case-competition teams: stress-test the team's issue tree, sanity-check market-sizing math, and sharpen slide action titles before presenting
What are this skill's strengths and limitations?
- 129 reference modules across 5 knowledge pillars give far more depth than a single prompt, spanning strategy, industry, commercial, and case-coaching knowledge
- Model-agnostic: works with Claude, Codex, Gemini, Copilot, Ollama, and any LLM via attaching SKILL.md
- Free and open-source (Apache 2.0 per README), with an explicit no-confidential-IP policy; forking and redistribution are permitted
- Ships directly usable tools: case scoring rubrics, step-by-step guesstimates, 100-day plans, SOW pricing models
- GitHub's license field shows NOASSERTION — verify the LICENSE file (README states Apache 2.0) yourself
- Purely knowledge-based: no scripts or automation, so output quality depends on the underlying model; models with limited context may not fit all 129 files in one session
- Industry benchmarks and the 2025-2026 demand archetypes are time-sensitive and need user validation for currency
- “MBB-grade” is descriptive marketing — it is not a substitute for real consultant judgment
How do you install this skill?
Option A (Claude Code plugin): run /plugin marketplace add DogInfantry/claude-skill-management-consultant-B1 followed by /plugin install management-consultant. Option B (manual): download skill/SKILL.md plus the entire skill/references/ folder and follow INSTALL.md in the repo. Other AI agents (Codex, Gemini, Copilot, Ollama) auto-load the skill via AGENTS.md, GEMINI.md, .github/copilot-instructions.md, and Modelfile respectively.
How do you use this skill?
After installation, start a new conversation and pose your problem or practice request directly, e.g. “Help me structure a profitability analysis for a $500M retail chain whose margins have been declining” or “Play the interviewer and run me through a McKinsey-style market entry case.” You can also trigger it with “Act as my management consultant.” The skill reads the relevant references/ files on demand before answering.