Execution reporting

Purpose

Execution without reporting is incomplete: the work happened, and nothing was learned from it. A report is not an activity log — it is reusable learning. What happened, where reality diverged from the plan, what that means, and what the next planning pass must know.

Over time reports are the only mechanism that moves guidance from plausible to validated. Every unvalidated threshold in this repository is waiting for a report that tested it.

When to use / when NOT to use

• Use after a meaningful burst of execution, at plan checkpoints, and when closing a work item.
• Do NOT write one for a trivial read-only lookup. Noise buries signal, and a repository full of empty reports is worse than one with few.
• Do NOT use a report as a progress tracker — that is the work item's job (durable-work).

Prerequisites