Fix weak coordination by designing explicit handoff protocols — specify what transfers, in what format, at what trigger — then practice three times

Why This Is a Rule

Weak coordination between agents typically manifests as handoff failure: Agent A finishes, Agent B should start, but the transition is vague — information is lost, the timing is unclear, or the format doesn't match what B needs. The fix isn't motivation or attention; it's protocol design: specifying the three elements that make any handoff reliable.

What information transfers: not "the results" but specifically "the prioritized task list with time estimates." Vague information specifications produce vague handoffs. In what format: not "whatever works" but "written list in the planning document, bullet points, with due dates." Format mismatches between producer and consumer are the most common handoff failure. At what trigger point: not "when it's ready" but "when I close the planning app and open the task manager." Observable triggers (Apply the camera test to triggers — if a camera can't detect the exact firing moment, the trigger is too vague) that mark the exact handoff moment.

The three deliberate practice executions after design serve the same function as The first five consecutive executions of a new trigger are non-negotiable — this is the window where automaticity lives or dies (first five executions for triggers): they install the handoff pattern through repetition before it needs to work automatically. Designing a protocol doesn't make it automatic; practice does.

When This Fires

• When coordination review (Four questions for 30-minute coordination reviews: who produced, did output reach consumers, what's the coordination ratio, where did agents interfere?) identifies handoff failures between specific agent pairs
• When output from one agent consistently fails to reach its intended consumer
• When a workflow's weak link is the transition between steps, not the steps themselves
• Complements Specify the information contract at every agent handoff — what exact output does the next step need from the current one? (information contracts) and Three-component handoff spec: output format, explicit expectations, and return protocol — ambiguous handoffs create bottlenecks (three-component handoffs) with the practice component

Common Failure Mode

Designing the handoff protocol but not practicing it: "I've specified what transfers and when — problem solved." Protocol design without practice produces a document, not a habit. The next three executions must deliberately follow the written protocol to build the procedural memory that makes the handoff automatic.

The Protocol

(1) Identify agent pairs with weak coordination (from Four questions for 30-minute coordination reviews: who produced, did output reach consumers, what's the coordination ratio, where did agents interfere? review or observed handoff failures). (2) For each weak pair, design the handoff protocol: What: list the specific information that must transfer from Agent A to Agent B. Format: specify the exact form (document, list, message, physical object). Trigger: define the observable event that initiates the handoff (Apply the camera test to triggers — if a camera can't detect the exact firing moment, the trigger is too vague). (3) Write the protocol down alongside the workflow documentation. (4) In the next three executions, deliberately follow the written protocol for this handoff. Don't rely on intuition — read the protocol and execute it as written. (5) After three deliberate executions: is the handoff reliable? If yes → the protocol is installed. If no → the protocol needs revision (information spec is wrong, format doesn't work, trigger is unclear). Revise and practice three more times.