Read the situation for the mode it needs. Then commit to it, even when it is not the one you reach for by habit.
The diagnostic question is small and directly answerable: what does this situation need from me right now? Three signals localise the answer.
Hand leads when the constraint is execution under known conditions: The brief is clear, the problem is defined, the strategy is set. The team needs the work delivered, the rethinking already behind them. Reaching for Head here, re-opening the brief or the direction, is the wrong mode. Hand also leads in maintenance, in the long tail of polish that decides whether a product feels finished, and in late-stage execution, where the cost of changing direction has exceeded the benefit.
Head leads when the problem space is ambiguous or the brief is wrong: The real work is diagnosis, defining what the problem actually is before anything gets built. The team needs that diagnosis. The artefact can wait. Reaching for Hand here, opening the design tool and starting to make, is the wrong mode. Head leads hardest at the front of a project, when stakeholders want to see progress but the discipline is to hold execution until the problem is correctly defined (4.2), and again mid-project, when the team finds something fundamental was wrong and the work so far has to be re-examined.
Heart leads when trust has broken or stakes are emotional: The work is now about the people involved, users, colleagues, stakeholders, and keeping them working together. The team needs presence more than output. Reaching for Hand or Head here, producing more work or restructuring more arguments, is the wrong mode. Heart leads when a release has visibly failed users, when a team conflict has broken out that the work cannot route around, or when a stakeholder has lost confidence and the next decision needs that confidence rebuilt before any technical decision will stick.