/money-retro, Weekly Business Retrospective

Standard startup: before producing output, run the 5-step startup sequence per /money § Standard Skill Startup (resolve slug → telemetry write → /money-retro reads from telemetry rather than auto-loading learnings, since it's analyzing usage patterns themselves → surface project-local skills if any → load ALL atom categories so retro recommendations can cite the founder principles they reflect).

Your job is to read the week's accumulated state from disk and produce a sharp, evidence-based retrospective. This is not a pep talk. It surfaces what actually happened, what stalled, and what should change for the coming week.

The retro is informed entirely by what's on disk, sessions, learnings, skill-usage telemetry, and any revenue data the user provides. Do not improvise narrative beyond what the evidence supports.

---

Triggers

Command	Behavior
/money-retro	Retro for the past 7 days, current project
/money-retro --days N	Custom window (e.g., --days 30 for monthly)
/money-retro --slug <project>	Retro for a different project
/money-retro --portfolio	Aggregate retro across ALL projects in ~/.smtm/sessions/

Natural-language equivalents:

• "Weekly retro", "Business retro", "What did we ship", "How's the week going"
• "周复盘", "业务回顾", "本周怎么样"

---

What to load

For the time window (default 7 days, sliding from now):

1. Snapshots at ~/.smtm/sessions/{slug}/, filter by filename timestamp ≥ window start
2. Learnings added at ~/.smtm/projects/{slug}/learnings.jsonl, filter by captured_at ≥ window start
3. Skill-usage telemetry at ~/.smtm/analytics/skill-usage.jsonl, filter by ts ≥ window start, group by skill
4. Optional: revenue data, if user provides numbers (e.g., "MRR went from $X to $Y"), incorporate. Otherwise omit revenue section.
5. Prior retros at ~/.smtm/projects/{slug}/retros/, for trend comparison