weekly-review

Weekly Review

Reference

• ID: S-ORG-13
• Category: Orchestration
• Inputs: artifacts/ directory, cogs/ai-ledger/, tasks/lessons.md, company.config.yaml
• Outputs: Formatted weekly summary (optionally saved to artifacts/decision-memos/)
• Used by: Orchestrator Agent, User (directly)
• Tool scripts: ./tools/ops/dashboard.sh, ./tools/security/posture-check.sh, ./tools/ops/token-ledger.sh

Purpose

Provide a weekly operating rhythm for solopreneurs and small teams. Surfaces what happened, what it cost, what's risky, and what to do next — all in one view.

Procedure

Step 1: Load Context

Read company.config.yaml for company name, stage, and persona names.

Step 2: What Shipped This Week

Scan all artifact directories for files modified in the last 7 days:

• List artifacts by type (PRD, RFC, QA report, etc.) with their current status
• Highlight any artifacts that were promoted (status changed to approved)
• Note any new artifacts created

Step 3: AI Spend

Read cogs/ai-ledger/entries.jsonl for entries in the last 7 days:

• Total input/output tokens and estimated cost
• Top 3 flows by cost (e.g., ship, prototype, code-review)
• Week-over-week trend if previous week data exists

Step 4: Support & Feedback Themes

Scan artifacts/decision-memos/ for CONV- prefixed artifacts from the last 7 days:

• Count of conversations logged
• Top pain points mentioned
• Any patterns emerging across conversations

Step 5: Security Health

Run ./tools/security/posture-check.sh --brief (if available):

• Current posture status (HEALTHY/DEGRADED/AT_RISK)
• Any new findings since last week

Step 6: Lessons Learned

Read tasks/lessons.md for entries added this week:

• New patterns or corrections captured
• Recurring issues to address

Step 7: Next Week Priorities

Based on the scan:

• Artifacts in draft/review that need attention
• Failing or incomplete gates
• Upcoming deadlines or commitments

Step 8: Output Format

## Weekly Review — [Company Name] — Week of [Date]

### What Shipped
- [list of completed/promoted artifacts]

### AI Spend
- Total: $X.XX (NK input, NK output tokens)
- Top flows: [list]

### Customer Signal
- [conversation count] conversations logged
- Top themes: [list]

### Security
- Posture: [status]
- [any new findings]

### Lessons
- [new lessons this week]

### Next Week
- [ ] [priority 1]
- [ ] [priority 2]
- [ ] [priority 3]

Step 9: Save (Optional)

If user passes --save, save to artifacts/decision-memos/REVIEW-WEEKLY-{YYYY-MM-DD}.md with appropriate frontmatter.

Quality Checklist

• [ ] All artifact directories scanned for recent changes
• [ ] AI spend calculated from ledger entries
• [ ] Security posture checked
• [ ] Priorities are actionable and specific
• [ ] Output is concise (fits in one screen)