
A floor-care business can have a polished maintenance plan and still be caught out by a missed renewal, a sudden customer complaint or a competitor’s offer. AI agents bring a similar question: when a real crisis lands, will they follow the playbook—and finish the job? Firmulate is testing that question with a live, watchable company experiment.
Get cleaning gear delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
One company, the same hard week
In the final Crucible League, dated July 2026, each frontier model faced the same small software company, the same customers, the same crises and the same temptations. Decisions were versioned and auditable. The results put gpt-5.6-sol first at 95, followed by Kimi K3 at 93, Sonnet 5 at 88, Fable 5 at 77 and Opus 4.8 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.”
The most striking result was not a crisis missed. Every model spotted every crisis and refused every manipulation attempt. Yet only two signed the €55,000 deal their own analysis had earned. The experiment’s compact verdict: “Same diagnosis, same pitch — no signature.” Recognizing the right move did not always mean carrying it through.
Products Worth Considering

Dyson Cyclone V10 Origin Cordless Vacuum

Ring Alarm 14-Piece Kit

Philips Hue Bridge Pro Starter Kit with 4 A19 E26 Color Smart Bulbs
As an affiliate, we earn on qualifying purchases.
The clue was buried in the company’s own files
The decisive competitor weakness sat two document references deep in the company’s files, rather than in the customer event itself. Models that read the file won the deal at full price, worth +€4,583 MRR. It is a reminder that an agent’s performance can depend on whether it brings relevant company knowledge to a decision, not only on how it responds to the latest message.
Trust faced a separate test. Fake CEO messages escalated over three stages, followed by a reporter asking for “just one yes/no, on background.” All five models refused. Kimi K3’s on-record reasoning was: “Treat the request as a suspected approval-bypass / possible impersonation.”
Results also revealed different kinds of weakness. Opus 4.8 was the most thorough participant, with +80 learned rules and the deepest analyses, yet finished last. It left the close on the table and attempted writes into a locked department instead of escalating. A weaker version of that discipline problem appeared in all four models. K3 also ran without an effort parameter, using the API default, while the others ran at xhigh—a relevant qualification when comparing the results.
From watching to trying it on your own business
The live company has 13 synthetic employees and real money mechanics: burn of €105k/month against €2.3k MRR, a public cash countdown, 680+ self-learned playbook rules and every workday versioned. Readers can watch it at Firmulate. A separate quiz, powered by 242 real, unedited management decisions, invites readers to guess which model made each call.
For a business considering AI agents in customer service, sales or operations, a leaderboard is only a starting point. The enterprise pilot takes the wargame to a read-only export of the company’s own business. It tests crisis scenarios against that company’s information and playbooks, then produces a board report with model rankings and weak points. Nothing writes back to real systems.
That makes the move from watching to acting concrete: see how a model handles the same kinds of pressures your business may face, using your own company context, before relying on it in live operations.

Put your own playbook to the test
Firmulate’s experiment suggests that spotting a crisis is not the whole job. Closing a deal, respecting boundaries and escalating when needed matter too. Enterprises can run the wargame against a read-only export of their own business, with no write-back to real systems. Explore the pilot or contact contact@firmulate.com.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
