FICTIONAL EXAMPLE — not a real person, not real corpus. Part of the examples/ worked corpus (persona: Sam Rivers). See examples/README.md.

Weekly Review — 2026-W09

~20 minutes. Cluster the week, check coherence, surface candidates. Led by Coach + Journalist. (This is the artifact a forker produces with the /weekly skill — shown here filled in so you can see the coherence check actually run.)

The week in facts

• Store page draft went up (private) on 2026-02-27; the hook line is still weak.
• Cut nothing from the tutorial yet — it's still too long.
• Zero movement, two late nights, no walks logged. Skipped the usual Tue/Thu runs.

Execution vs intent

Stated priority was the demo funnel, and the project did move (page draft) — so intent and execution matched on the work axis. But the move was small and the load-bearing next action (cut the tutorial to 90 seconds) didn't happen. The week went to the easy 80% (a page) and dodged the hard 20% (the tutorial cut that real players actually need).

Coherence report

Run the six checks from steering/coherence.md. Findings are questions, not edits.

• Orphans: None. project:build-demo-and-wishlist-page → goal:launch-paid-game-v1 → vision + principle:ship-before-its-ready. The ladder is intact end to end.
• Contradictions: None stated, but a watch: the launch goal's timeline is tightening while the vision's load-bearing condition is "rested." Not a contradiction in the files yet — a pressure to name before it becomes one.
• Starved areas: area:health is starved. Zero evidence or action this week, two late nights, runs skipped — while work got the week's only reps. The health standard ("rested enough to sustain the pace for years") is going unfunded.
• Over-capacity: No. One active project, one active goal — capacity is fine; if anything there's room.
• Stale: None. The project's last_moved is 2026-02-27 (this week).
• Drift (revealed/aspirational): Healthy. The one principle is revealed and has fresh, dated evidence; nothing aspirational is masquerading as lived.

Candidates (you decide)

Clearly-labeled candidates the agent surfaced. Accept, rewrite, or reject — none are committed here.

• Candidate habit: "Runs are a default, not a reward — Tue/Thu before work, blocked on the calendar." (Surfaced from the starvation finding; you decide if it's real.)
• Candidate principle (aspirational): "I fund my health like a load-bearing area, not the slack I spend first under deadline." Marked aspirational — there's no evidence yet that Sam lives this. It earns revealed only when stories back it.

The one question

Am I quietly trading the "rested" half of the vision to hit the launch date — and if so, is that a decision I'm making on purpose, or drift I haven't noticed?

Commitments for next week

• Cut the tutorial to 90 seconds and record one stranger playing it (the real next action, not the easy one).
• Two runs, calendar-blocked, before the funnel work — to test the candidate habit.
• Carry the one question into the monthly review; don't resolve it on impulse here.