What problem does it solve? Personal weekly reflection across multiple projects, notes, and communication channels is hard to do honestly: activity gets inflated into outcomes, gaps get filled with guesses, and private sources get overexposed. This Skill turns one bounded week of authorized work and knowledge evidence into a concise, evidence-backed review that keeps outcomes, activity, carryover, lessons, and next-week focus strictly separate. ## Core Features & Use Cases - Evidence-bounded synthesis: Builds a normalized activity ledger from authorized sources, deduplicates by stable origin IDs, and classifies items as outcomes, meaningful activity, decisions, carryover, or lessons without upgrading them. - Privacy and profile governance: Enforces a disclosure ceiling (private-only, internal, outward-safe), resolves worklog and personal profiles only through explicit locators, and never reproduces private paths, tags, or identities. - Calibrated reporting: Classifies source coverage as complete, partial, stale, inaccessible, or missing, and returns an explicit insufficient-evidence result instead of filler when evidence is sparse. - Use Case: On Monday morning, ask for a review of the previous week using your configured worklog profile; receive destination-neutral Markdown with outcomes, carryover, friction patterns, and at most three ranked next-week focus items, each linked to evidence. ## Quick Start Ask the AI to run a weekly review for the previous week in your timezone using your authorized worklog profile and private-only privacy scope.
