Flashpoint

A policy survives contact with a spreadsheet. It does not always survive contact with a Tuesday afternoon, a changing room, two upset people and a manager who has thirty seconds to say something. The gap between what a policy says and what happens in that moment is where almost all the damage actually occurs.

What it does. Flashpoint is what takes the scenario library off the page. The Toolkit holds a library of scenarios and a library of personas; Flashpoint walks you through one of those situations step by step, in real time, with the personas as the other people in the room — digital twins acting out the drama around you rather than case studies you read about.

You take a role — the duty manager, the HR lead, the head of comms, the on-call clinician — and work the situation from inside that seat. You are not playing a named person; you are standing in the role, while the cast reacts around you. Each case is short and self-contained: something has gone wrong, or is about to, and colleagues, families, journalists and social media are reacting as it unfolds.

The tool is identity-aware, not identity-led. In some cases you will know a person's identity, because they have told you, it is evident, or it is genuinely relevant to your duty of care. In others you will not, and asking would be wrong. Either way the test is the same — handle it so the decision is defensible whether or not you know, never making someone's identity the basis of the call and never demanding a disclosure you do not need. The dilemma itself is always described in full.

A case unfolds as a sequence of beats, and each beat is a decision: how do you interject, or how do you respond, right now? Not what the policy says in principle — what you actually say, in this room, to this person, while everyone is still watching. The question the tool keeps putting to you is simply how will you do?

There is no time pressure on the click, because the pressure belongs in the situation rather than the interface. Six consequence meters move as you decide, four of them better when they rise: dignity, privacy, trust and operational control. No single beat wins or loses it — a good call late does not undo a careless one early, and that is the point. At the end you are asked to explain what you think actually happened and to commit to a verdict.

What you get. A full debrief and coaching on how you actually handled it — not just a score. You reconstruct what made the harm avoidable, with suggested scripts for the moments you found hardest and specific points to watch next time. The culprit is always a decision or a process, never a person and never anyone's identity, and what surfaces is usually not a rule that was missing but a step in your own process that quietly forced somebody's hand.

That is the part that transfers. A walkthrough you cannot remember afterwards changes nothing; wording you can actually use, and a short list of what to watch for, is what turns a rehearsal into practice.

Who uses it. Line managers rehearsing the moment they dread. HR and EDI leads testing whether a policy is survivable in practice. Teams who would rather find out in rehearsal.

Where it fits. The Preventative Duty check surfaces the need. What Flashpoint exposes tends to route into Complaints Triage and Challenge Response, and the process failures it uncovers belong in front of Defensibility.