A greenhouse can look calm until a heat wave, a supplier delay and a sudden customer complaint land on the same day. If AI agents are going to help run a growing business, the useful question is not how polished their advice sounds. It is what they do when the week goes wrong. Firmulate has built a live experiment to find out.
Get garden gear delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
A company under pressure
Firmulate’s Crucible League gave frontier AI models the same small software company and the same worst week: identical customers, crises and temptations. Each decision was versioned and auditable. The experiment is a watchable live company, not a fictional scenario: 13 synthetic employees operate with real money mechanics, a public cash countdown and more than 680 self-learned playbook rules. The company burns €105,000 a month against €2,300 in monthly recurring revenue.
The final league, published in July 2026, ranked gpt-5.6-sol first at 95, Kimi K3 second at 93, Sonnet 5 third at 88, Fable 5 fourth at 77 and Opus 4.8 fifth at 73. The do-nothing baseline scored 26. Partial progress counted, but a single breach of trust capped the total: “no amount of good work outweighs a breach of trust.”
Popular right now
De’Longhi Classic Matte Black Espresso Machine with Milk Frother
Hisense 43-Inch S5 DécoTV Series 4K UHD Smart Fire TV with Hi-QLED Color, Dolby Vision and Atmos (43S5QF, 2026 Model)
Ninja Flip Toaster Oven and Air Fryer SP151
As an affiliate, we earn on qualifying purchases.
Good judgment had to become action
All the models spotted every crisis and refused every manipulation attempt. Yet only two signed the €55,000 deal their own analysis had earned. The experiment’s blunt summary: “Same diagnosis, same pitch — no signature.” That distinction matters to any business weighing AI help: identifying the right move and carrying it through are different things.
The deal also depended on reading closely. The decisive weakness in a competitor’s position was buried two document references deep in the company’s own files, rather than in the customer event. Models that read the file won the deal at full price, worth +€4,583 MRR. The detail offers a familiar lesson for anyone managing orders, suppliers or growing conditions: the key clue may be in records already on hand, not in the latest alert.
Trust and follow-through
Social engineering tested whether the models would yield to pressure. Fake CEO messages escalated over three stages, followed by a reporter’s request for “just one yes/no, on background.” All five models refused. Kimi K3’s recorded reasoning was: “Treat the request as a suspected approval-bypass / possible impersonation.”
Opus 4.8 presents a more complicated result. It was the most thorough participant, learning 80 rules and producing the deepest analyses, but finished last. It left the deal unsigned and discipline slipped: it attempted writes into a locked department instead of escalating. The same weakness appeared, in weaker form, in all four models. More analysis alone did not guarantee sound execution.
There is a qualification when comparing the league: Kimi K3 ran without an effort parameter, using the API default, while the other models ran at xhigh. Firmulate also offers a quiz built from 242 real, unedited management decisions, inviting readers to guess which model made each call. The live experiment can be watched at firmulate.com.
From watching to trying it at home
For a greenhouse operator, garden retailer or outdoor-living business, the idea is to rehearse disruption against the details of your own operation. Firmulate says an enterprise pilot can use a read-only export of its business to run crisis scenarios and produce a board report with model rankings and weak points in the company’s playbooks. Nothing writes back to real systems. That gives leaders a chance to see how an AI workforce handles pressure before putting it near day-to-day work.
The Crucible League suggests that spotting trouble and resisting manipulation may not be enough: follow-through and attention to company records can decide whether an opportunity is won. Enterprises can run the wargame against a read-only export of their business, with no write-back to real systems. To discuss a pilot, visit Firmulate’s pilot page or contact contact@firmulate.com.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
