Case studies
A standard reads differently from every chair. These are complete projects taken end to end through FitSD — intake, both gates, build, Service Acceptance, live operation, succession and the governance retro — with every participant logging where the standard helped, where it was silent, and where it got in the way.
They are non-normative: evidence and test tooling, not part of the standard.
Runs
A run is one scenario, at one organisation, conducted per the method, against one version of the standard. Re-running the same scenario against a later version is how you see whether the standard actually improved.
| Run | Scenario | Organisation | Standard | Seats | Findings |
|---|---|---|---|---|---|
| 2026-08 | CertHub | Fernbrook Software | v0.3 | AI simulation | 110 |
Building blocks
Each piece is reusable on its own: set a different scenario at the same organisation, re-run a scenario against a later version, or run any of them by the same method.
Scenarios
The work pushed through the standard. Names no version, so it can be re-run.
- CertHub —
Organisations
The simulated company a scenario is set in. Version-neutral, so the same organisation can host many scenarios and be run against any version.
- Fernbrook Software () —
Run one yourself
The suite is a public repository and it is meant to be picked up. Bring your own scenario, run it against whichever version of the standard you care about, and contribute the result back — it appears here automatically. Runs with people in the seats are more valuable than AI ones, not less: an all-AI run shares its blind spots across every seat.
How to contribute a run ·the AI harness ·the protocol
Method
How a run is conducted. Scenario-agnostic, shared by everything.
How a run works ·Protocol ·Findings template · 13 role briefs, one per stage-seat.