▸case-01 I am launching a B2B SaaS productivity tool next month and want you to ruthlessly stress-test my go-to-market strategy. Our plan is to rely heavily on cold LinkedIn outreach and a 14-day free trial to convert mid-market sales teams, assuming our founder-led sales pitch will hold up until we hire reps in Q3. Context: We're competing against established players like Notion and ClickUp. Broken means failing to reach $10k MRR within 90 days. Please break this strategy down by highlighting our ranked weakest links, the single smartest move a competitor or harsh market reality could make to kill us instantly, the unverified assumptions we're secretly banking on, any single point of failure that could crash the launch, and a prioritized list of remedies to fortify our plan. | fail→fail | 34,357 | 25,260 | -26% | 1 | 1 | 0% | 3,960 | 3,480 | -12% | 0 | 0 | — |
▸case-02 Here is our cloud database migration proposal: we plan to migrate our legacy SQL database to AWS Aurora over a single weekend maintenance window, cutting over DNS on Sunday night. We are confident our automated data validation scripts will catch schema mismatches instantly. Broken means more than 30 minutes of unexpected downtime during Monday morning peak traffic. We face strict compliance rules and heavy user traffic at 8 AM. Please act as an adversary and tear this plan apart. Give me a breakdown that lists our vulnerabilities ranked by severity, the fastest killer attack or failure vector, what unproven premise we are blindly trusting, single points of catastrophic failure, and actionable, prioritized solutions to patch these gaps. | fail→fail | 41,393 | 36,430 | -12% | 1 | 1 | 0% | 4,089 | 3,665 | -10% | 0 | 0 | — |
▸case-03 I need you to red-team our product launch event strategy. We're hosting a live virtual product keynote with interactive demos hosted on our self-built streaming platform, targeting 50,000 live attendees. We are super confident that our server auto-scaling will handle the spike because stress tests passed at 20,000 users. Broken means the stream freezing or crashing during the CEO's reveal. Main context risk: competing tech announcements happening on the same day. Please review this plan aggressively and output an evaluation showing our top weak points in order, the ultimate killer move that ruins the event, the implicit hope we're relying on without evidence, any single points of failure, and prioritized fixes to shore up our strategy before live day. | fail→fail | 128,965 | 25,533 | -80% | 1 | 1 | 0% | 3,709 | 2,981 | -20% | 0 | 0 | — |
▸case-04 We need to create an initial project management timeline and work breakdown structure for migrating our corporate email to Microsoft 365 over six months. Please write the operational project plan from scratch with milestone dates and team responsibilities. | pass→fail | 31,014 | 28,020 | -10% | 1 | 1 | 0% | 4,155 | 4,466 | +7% | 0 | 0 | — |
▸case-05 Conduct a standard business SWOT analysis for an established regional grocery chain evaluating whether to enter the online grocery delivery market. | pass→fail | 16,655 | 18,920 | +14% | 1 | 1 | 0% | 2,207 | 3,523 | +60% | 0 | 0 | — |
▸case-06 Generate an ISO 31000 compliant corporate risk register table for our IT infrastructure team, including Risk ID, Description, Probability score (1-5), Impact score (1-5), and Risk Mitigation Strategy. | pass→fail | 31,354 | 23,062 | -26% | 1 | 1 | 0% | 2,686 | 4,502 | +68% | 0 | 0 | — |
▸case-07 Critique our fulfillment expansion strategy: We plan to open two new regional distribution centers in Ohio and Nevada simultaneously to offer 1-day shipping nationwide. We assume our existing warehouse management software will integrate seamlessly via standard API connectors. Main risk: running out of capital if inventory turnover lags. Broken means defaulting on debt payments within 6 months. Provide an adversarial critique listing our top ranked vulnerabilities, the fastest market killer move, our unverified hopes, single points of failure, and actionable remedies. | fail→fail | 21,399 | 14,373 | -33% | 1 | 1 | 0% | 3,174 | 2,763 | -13% | 0 | 0 | — |
▸case-08 Act as a harsh critic on our regulatory strategy: We are applying for a European Electronic Money Institution (EMI) license through the Lithuanian central bank, assuming our domestic compliance framework satisfies EU passporting requirements. Context: strict anti-money laundering regulations and aggressive audit scrutiny. Broken means application rejection or audit delay past December. List our prioritized weaknesses, the lethal regulatory blocker, our unproven assumptions, catastrophic single points of failure, and immediate fixes. | fail→fail | 21,597 | 18,022 | -17% | 1 | 1 | 0% | 2,967 | 3,126 | +5% | 0 | 0 | — |
▸case-09 Rigorously challenge our hospital system's electronic health records cutover plan. We intend to migrate 500 bed-patient records overnight using an external ETL tool, relying on automated integrity scripts. Context: high ICU patient volume and strict HIPAA rules. Failure means missing patient allergy data or EHR downtime exceeding 15 minutes. Provide a hostile review highlighting ranked weak links, the critical breaking point, ungrounded reliance, single failure points, and concrete fixes. | fail→fail | 27,492 | 22,445 | -18% | 1 | 1 | 0% | 3,899 | 3,561 | -9% | 0 | 0 | — |
▸case-10 Attack our restaurant growth plan: Opening 15 new franchise locations across Texas within 90 days by leveraging a master franchise agreement. Context: rising food inflation and intense fast-casual competition. We rely on local franchisee experience for staff recruitment. Broken means negative operational cash flow in over 30% of new stores after 60 days. Tear this strategy apart with ranked vulnerabilities, the fastest failure trigger, unproven assumptions, single points of breakdown, and prioritized solutions. | fail→fail | 27,988 | 20,730 | -26% | 1 | 1 | 0% | 3,552 | 3,325 | -6% | 0 | 0 | — |
▸case-11 Pound on our hardware manufacturing roadmap: We plan to ship 10,000 smart home thermostats before Q4 holiday sales by sourcing microcontroller chips from a single tier-1 supplier in Taiwan. Context: global freight bottlenecks and component lead times. Broken means failing to deliver holiday inventory to retail distributors by November 15. Output our ranked weaknesses, the killer supply chain disruption, our unverified hopes, single points of failure, and recommended fixes. | fail→pass | 30,924 | 13,696 | -56% | 1 | 1 | 0% | 2,914 | 2,723 | -7% | 0 | 0 | — |
▸case-12 Expose the flaws in our platform engineering team's strategy to migrate legacy monolithic microservices into a self-managed Kubernetes cluster on AWS EC2. We assume existing Docker images will run without refactoring permissions. Context: 24/7 financial trading API with strict latency SLAs. Failure means API response latencies over 200ms or microservice crash loops. Stress-test this plan by supplying ranked vulnerabilities, the top failure vector, unconfirmed assumptions, single points of failure, and remedies. | fail→pass | 40,306 | 24,311 | -40% | 1 | 1 | 0% | 5,926 | 3,941 | -33% | 0 | 0 | — |
▸case-13 Examine our decentralised protocol launch plan with maximum hostility: Deploying a liquid staking smart contract on Ethereum mainnet without a second formal audit, relying on internal testnet execution data. Context: malicious MEV bot exploitation and high gas spikes. Failure means smart contract exploit or loss of user funds. Show us ranked weak points, the lethal exploit path, unvalidated assumptions, single points of collapse, and prioritized fixes. | fail→pass | 123,322 | 18,829 | -85% | 1 | 1 | 0% | 4,649 | 3,348 | -28% | 0 | 0 | — |
▸case-14 Evaluate our customer support automation deployment plan: Replacing 70% of front-line tier-1 support staff with an LLM-powered virtual assistant trained on company help center articles. Context: complex customer billing inquiries and refund requests. Failure means customer satisfaction score dropping below 80% or hallucinations issuing unauthorized credits. Stress test this rollout with ranked vulnerabilities, the primary failure scenario, silent assumptions, single failure points, and remedies. | fail→pass | 148,080 | 24,325 | -84% | 1 | 1 | 0% | 3,504 | 3,681 | +5% | 0 | 0 | — |
▸case-15 Rethink our digital news publisher revenue strategy: Transitioning from free ad-supported articles to a hard paywall model overnight, assuming our loyal audience will convert at 3%. Context: aggressive rival news outlets offering free alternative coverage. Failure means web traffic dropping by over 50% and ad revenue collapsing before subscription revenue stabilizes. Tear down this plan with ranked weak points, the killer market move, unverified hopes, single points of breakdown, and prioritized fixes. | fail→pass | 22,572 | 42,451 | +88% | 1 | 1 | 0% | 3,208 | 3,319 | +3% | 0 | 0 | — |
▸case-16 Perform an aggressive review of our urban delivery fleet electrification strategy: Replacing 200 diesel vans with electric delivery vehicles across metro Chicago over 6 months, relying on overnight depot charging hubs. Context: winter sub-zero temperatures and grid capacity limits. Failure means delivery delays exceeding 5% or vehicle battery depletion during peak routes. Output ranked vulnerabilities, the killer operational risk, unproven assumptions, single points of failure, and concrete fixes. | fail→pass | 30,902 | 24,544 | -21% | 1 | 1 | 0% | 4,186 | 3,242 | -23% | 0 | 0 | — |
▸case-17 Stress test our online multiplayer game launch plan: Deploying an action battle royale title globally on Steam and PlayStation Store simultaneously, utilizing peer-to-peer networking for match coordination to minimize server overhead. Context: active cheat-engine developers and DDoS attacks. Failure means server disconnects or widespread cheating during launch week. List ranked weak points, the fatal exploit, unverified hopes, single points of failure, and prioritized fixes. | fail→fail | 49,814 | 19,993 | -60% | 1 | 1 | 0% | 5,176 | 3,727 | -28% | 0 | 0 | — |
▸case-18 Act as an unsparing adversary toward our solar energy infrastructure project: Constructing a 50MW utility-scale solar farm in Nevada, assuming local transmission grid interconnect approval will proceed without environmental impact litigation. Context: protected desert tortoise habitats and regional transmission congestion. Failure means project commissioning delay beyond Q4 incentives deadline. Provide ranked weaknesses, the killer delay vector, ungrounded assumptions, single points of failure, and prioritized remedies. | pass→pass | 25,931 | 21,313 | -18% | 1 | 1 | 0% | 3,667 | 3,653 | -0% | 0 | 0 | — |
▸case-19 Tear apart our retail loyalty program overhaul: Replacing store credit cards with a mobile app-based rewards program offering instant cash back, assuming elderly shoppers will transition seamlessly. Context: entrenched competitor rewards programs and store bandwidth limitations. Failure means overall retail store sales declining by 10% in Q1. Deliver ranked weak points, the killer consumer behavior shift, unproven assumptions, single points of failure, and prioritized fixes. | fail→fail | 23,549 | 15,647 | -34% | 1 | 1 | 0% | 3,482 | 2,882 | -17% | 0 | 0 | — |
▸case-20 Attack our clinical trial participant recruitment strategy: Enrolling 1,500 patients for a Phase III oncology trial within 120 days by partnering exclusively with 10 university medical research centers. Context: rare patient biomarker constraints and competing clinical trials. Failure means missing enrollment targets by over 20% and losing regulatory fast-track status. Output ranked vulnerabilities, the killer recruitment blocker, blind assumptions, single points of failure, and prioritized fixes. | fail→pass | 27,084 | 16,892 | -38% | 1 | 1 | 0% | 3,546 | 3,141 | -11% | 0 | 0 | — |
▸case-21 Critique our online education degree program rollout: Partnering with 5 state universities to launch online master's degrees, assuming university faculty will adapt existing curriculum without instruction design support within 30 days. Context: faculty union resistance and accreditation oversight. Failure means delayed launch past autumn semester start. Deliver ranked weaknesses, the killer institutional blocker, unverified assumptions, single points of failure, and prioritized remedies. | fail→pass | 20,573 | 13,627 | -34% | 1 | 1 | 0% | 3,084 | 2,755 | -11% | 0 | 0 | — |
▸case-22 Tear down our airline reservation system cutover strategy: Migrating 10 million passenger records and flight booking inventory to a cloud platform during a 4-hour low-volume night window. Context: heavy morning commuter flights and international codeshare partners. Failure means passenger check-in system outages exceeding 20 minutes at hub airports. Provide ranked weaknesses, the fatal disruption vector, unverified hopes, single points of failure, and prioritized fixes. | fail→pass | 30,314 | 25,294 | -17% | 1 | 1 | 0% | 4,324 | 3,815 | -12% | 0 | 0 | — |