10 — CASCADE REPORT (1:1)

Final step. Reads all cascade outputs and produces a Kathryn-facing report surfacing what needs her eyes before any client-facing artifact (recap email, prep email) ships.

Why this step exists

The 1:1 cascade auto-advances through 9 steps without review gates between them. One bad interpretation at any step can propagate. This step is the safety net — it concentrates everything Kathryn needs to review into one report.

This is NOT a gate (the cascade still completed). It IS a structured audit of what was produced.

When this step runs

Final step of the cascade. Auto-advances from /extract-lessons-learned (file 09).

Outputs (TWO)

What to scan

• The recap .md (from file 05)
• Gmail draft (recap email sent to client)
• Gmail draft (prep email for next session)
• Calendar events created
• Member CPM (cpm/[name]-cpm-[date].md)
• Drafted deliverables (drafts/) and action-items index
• Updated open-actions list
• Lessons-learned files (newly created or updated this run)

Cascade report template

# Cascade Report — [Member Name] 1:1, [YYYY-MM-DD]

**Member:** [Name]
**Session date:** [YYYY-MM-DD]
**Cascade completed:** [timestamp]
**Pre-send status:** [🟢 GREEN / 🟡 YELLOW / 🔴 RED]

---

## TL;DR

[One paragraph — what ran, what's ready to send, what needs your eyes before it ships.]

---

## Ready to send / ship (no concerns)

- [Output] — [path / Gmail draft subject]
- [...]

---

## Needs your eyes before sending

### 1. [Specific concern]
- **What it is:** [file path]
- **Why flagged:** [specific concern, not generic]
- **Specific question to resolve:** [what answer would clear this]

### 2. [...]

---

## Transcript ambiguities (if any)

[Verbatim quote, candidate readings, which the cascade went with, what would resolve.]

---

## Member CPM — what changed

- **New constraints this session:** [list or "none"]
- **Status changes:** [what moved, in which direction]
- **Conversion read:** [Stay TPC / Sprint candidate / etc.] — [new this session / unchanged]

---

## Drafted deliverables (action items)

| Deliverable | Path | Confidence |
|---|---|---|
| [name] | drafts/[file] | [solid / needs review] |

---

## Lessons captured for PBOS-General

| Lesson | New or appended? |
|---|---|
| [Name] | [New / +1 evidence row] |

---

## Kit revisions this run (the feedback loop)

**This section is the source of truth for every kit change made during this cascade.** Every error Kathryn catches becomes a row here; every row maps to a corresponding change-log entry in kit file 04 or 05. The report and the kit are bound together — neither updates without the other.

| # | Error caught | Where caught | Pass | Kit change | Change shape | File + section | Status |
|---|---|---|---|---|---|---|---|
| 1 | [Specific failure] | [Pre-send Kathryn review / mid-cascade / post-send] | [Pass 1 / etc.] | [What was added/changed] | [structural / row-stack / vocabulary / golden / loop] | [file + section] | [Applied + logged / Pending] |
| 2 | [...] | [...] | [...] | [...] | [...] | [...] | [...] |

**Change shape values:**
- `structural` — change to file 05 step ordering, new step inserted, ledger category added (the most powerful kind of fix; addresses the failure upstream of where it surfaced)
- `vocabulary` — file 02 terminology lock (forbidden term, cadence vocabulary, sign-off rule)
- `golden` — file 03 example revision (positive or negative space)
- `loop` — file 10 cascade-report mechanism change
- `row-stack` — new row in file 04 Blocking Failures or Common Failure Modes (the *least* powerful kind of fix — use only when the failure is genuinely a post-draft check, not a structural gap)

**Pattern signal — escalate when reactive patching keeps stacking rows.** If three consecutive cascade runs keep producing `row-stack` fixes for the same shape of error, that's a structural failure trying to surface as a checklist item. Escalate to a `structural` change (new step, ledger category, file ordering) before file 04 grows another row.

**Empty if no errors caught this run.** Each row maps 1:1 to a change-log entry in file 04 or 05.

### Loop closure check (REQUIRED)

Before this report ships:
- [ ] Every row above has a matching entry in kit file 04 or 05's change log
- [ ] Every kit change-log entry from this run is listed above
- [ ] Vault commit hash referenced in the report and in the change log
- [ ] If multiple QC passes happened (Kathryn caught more on review), each pass gets its own batch of rows — don't collapse them

---

## Pre-send checklist

Before recap email goes to [Member]:
- [ ] Recap voice matches prior emails
- [ ] No stakeholder critique that would feel awkward if forwarded
- [ ] Action items match what was actually committed
- [ ] [Other situation-specific checks]

Before prep email sends 2 business days before next session:
- [ ] Prep email voice matches the recap
- [ ] Action items list is current (matches updated open-actions)
- [ ] [Situation-specific checks]

---

*Cascade report · [YYYY-MM-DD] · Generated by /cascade-report-1on1*

Pre-send status — color rules

Gmail draft (output #2)

• To: kathryn@creatingyourplan.com
• Subject: Pre-send review — [first name] 1:1 [M-DD] ([N] items need your eyes before sending)
• Or: Pre-send review — [first name] 1:1 [M-DD] (all clear — ready to send) if green
• Body: HTML version of the markdown report, inbox-readable, no workflow jargon
• Plain English in the subject — no "cascade," no internal terminology

Quality gate

• TL;DR is one paragraph, not five
• "Needs your eyes" items have specific questions, not generic flags
• Pre-send status color is defensible — if anything is non-trivially uncertain, NOT green
• Pre-send checklist is situation-specific (not boilerplate)
• Kit Revisions section is populated for any error Kathryn caught this run — and every row has a matching change-log entry in kit file 04 or 05. Empty section means no errors caught (not "I forgot to fill it in")
• When Kathryn catches MORE errors on her review of this report, the report gets a new pass section (Pass 2, Pass 3...) — never overwrite the prior pass. Each pass = own batch of Kit Revisions rows + own batch of change-log entries in the kit. Audit trail intact.

End of cascade

This is the final step. When complete, present to Kathryn:

"Cascade report delivered. [N] items flagged for your review. Pre-send status: [GREEN/YELLOW/RED]. Report at members/[name]/sessions/[name]-cascade-report-[date].md and Gmail inbox draft."