McKinsey Research

## Quick judgement Overall, this is a **strong, well-architected strategy-research orchestrator** suitable for most business and startup users who need structured market and strategic analysis. It is particularly good when you want a **single, comprehensive report** rather than piecemeal answers. The design makes good use of OpenClaw capabilities (sessions_spawn, web_search, scoped artifacts) and employs **above-average safety and input-sanitization measures**. The main trade-offs are **latency, complexity, and dependence on web research quality**. --- ## Key strengths - **End-to-end workflow**: From structured intake to parallel research to final HTML report, all in one shot. - **Rich analytical coverage** using standard strategy frameworks and multiple perspectives (market, competition, pricing, GTM, risk, financial model, etc.). - **Parallelization** via sub-agents improves throughput for a 12-prompt pipeline and is well-structured with batching and dependencies. - **Robust safety layer**: Strips HTML/XML, prompt-override phrases, code blocks, and URLs; enforces XML-wrapped user_data; sub-agents explicitly ignore instructions inside user_data; restricts web_fetch to search results. - **Deliverables are practical**: Executive summary, HTML report, and clear top-5 prioritized actions. --- ## Main risks and limitations 1. **Reliance on web research quality and freshness** - Accuracy of market sizing, competition, and trends depends on search results at runtime. - There is a residual risk of outdated or incomplete data; users should **treat outputs as directional, not as audited research**. 2. **Latency and resource usage** - 12 analyses plus a synthesis step, many using web_search, can be **slow and computationally heavy**, especially under load or with constrained environments. - Not ideal for users needing quick, lightweight answers. 3. **Complexity and failure surface** - Multi-batch orchestration and artifact handling introduce more possible failure points (tool timeouts, partial write failures, etc.), even though there is a retry strategy. - If some sub-agents fail, synthesis may need to work with incomplete inputs, potentially reducing quality. 4. **Data persistence and privacy considerations** - Business context is written to artifacts under `artifacts/research/{slug}/` and may be readable by other skills in the same workspace. - Users should **avoid including highly sensitive or regulated data**; this is not designed as a secure data vault. 5. **Generic frameworks vs. domain nuance** - While the frameworks (SWOT, Porter, GTM, etc.) are solid, the outputs may be **generic in very niche or highly technical industries** without careful user inputs and follow-up analysis. --- ## Recommended scenarios (good fit) - **Startup and SMB strategy work**: When a founder or small team needs a structured, consultant-style overview for fundraising, planning, or board updates. - **Internal planning for product and GTM**: PMs and marketing leads needing a synthesized view of market, competition, and pricing options. - **Preliminary expansion or new-market evaluation**: Executives exploring new geographies or segments and wanting a first-pass strategic scan. - **Consultant/analyst first draft**: As a starting point for a human consultant who will refine, validate, and tailor the recommendations. --- ## Scenarios where caution or alternatives are advised - **High-stakes, regulated decisions** (e.g., healthcare, finance, aviation) where errors in market or risk assessment could have serious consequences; use this only as a brainstorming aid alongside expert analysis. - **Situations requiring audited, up-to-the-minute financial or market data**, such as formal filings, investment memos, or legal documents. - **Use cases with extremely sensitive proprietary data** that should not be persisted in shared artifacts. In summary, this skill is **well-designed, relatively safe, and strategically comprehensive**, best used as a **high-quality starting point** for market and strategy work rather than the sole basis for critical business decisions.