Closed-Loop Analytics Upgrade

Principle

A workflow is not a closed loop until it checks whether the change worked and updates the playbook from that evidence.

For marketing skills, that means pulling analytics after the change window. Manual opinions are useful. Platform truth wins.

Core pattern

1. Input: usage logs, recommendations, shipped changes, platform analytics, owner feedback, and cost/runtime data.
2. AI action: compare baseline vs candidate, find repeatable signals, and propose a skill/playbook patch.
3. Output: candidate patch to a prompt, skill, connector, brief template, scoring rubric, or next-action rule.
4. Judgment: success rate, speed, cost, quality, human correction rate, and actual performance delta.
5. Self-improvement: promote only if it beats baseline. Otherwise keep testing, rollback, or mark unproven.

Analytics by surface