← ClaudeAtlas

weekly-reviewlisted

Reusable skeleton playbook. Walks the operator through a weekly review across bets, offers, pushes, playbook runs, outcomes, decisions, logs, and checkpoints. Not yet a single executable orchestrator.
noontide-co/mainbranch · ★ 27 · AI & Automation · score 81
Install: claude install-skill noontide-co/mainbranch
# weekly-review (skeleton) This is a reusable operating recipe, not a one-off weekly review file. Use it to surface what changed, decide what needs attention, record outcomes from active pushes and playbook runs, promote approved research or decisions, and save a checkpoint. **Current status: skeleton.** This file documents the intended review chain and artifact routing. It is not yet executable as one coherent workflow with a single progress surface, automatic diffs, or cross-skill data passing. Run the checks and skills manually today. ## Playbook Versus Run This file is the reusable engine recipe under `.claude/playbooks/weekly-review/`. The concrete review belongs in the business repo. Use: - `log/weekly-review-YYYY-MM-DD.md` for the week-level chronology, decisions, blockers, and handoff notes; - `pushes/<push>/playbooks/<playbook>.md` when reviewing one push playbook run's approvals, forks, validation evidence, manual steps, or linked outcomes; - `pushes/<push>/review-log.md`, `pushes/<push>/reviews/`, or outcome/log links when the review changes the push state; - `research/`, `core/`, `decisions/`, or `bets/` only when the operator approves durable memory updates. Do not write new work under `campaigns/`; that folder is legacy compatibility read only. ## Intended Flow ```text 1. /mb-start or mb status --json --peek Read repo health, updates, provider readiness, active bets, pushes, and recent activity. 2. mb validate Confirm frontmatter and lin