closure-retrospectivelisted

# Closure Retrospective ## Core Principle Use completed-task evidence to improve the context that shapes future decisions. Determine how the current context influenced the run, define the desired future context at the same decision point, and route the smallest durable change to the earliest semantic owner. Optimize decision quality and system coherence, not artifact count. Treat context as everything the worker can legitimately use at the decision point: governing facts, boundaries, judgment criteria, authority, feedback, and executable structure. Do not equate context with prose. Treat an observed outcome as evidence, not as a rule. Raw trajectory material and agent self-reports are low-trust telemetry. Keep them attached to the run, corroborate them against diffs, tool output, checks, review, runtime evidence, and accepted outcomes, and never inject them directly into future context. ## Closure Gate Run this skill only after the substantive task is complete or clearly entering final handoff. If implementation, debugging, research, product intent, architecture, or the acceptance boundary remains unresolved, finish or escalate that work first. Drop trivial work, one-off preferences, temporary debugging notes, unsupported interpretations, and findings whose only evidence is an agent self-report. ## Workflow 1. **Build an evidence packet.** Record the promised outcome, observable trajectory, intervention or decision, result, proof, human steering, and accepted or reje