code-ai-installer 3.0.0 → 4.0.1-a

package/domains/product/locales/en/.agents/skills/session-prompt-generator/SKILL.md

@@ -1,450 +1,474 @@
----
-name: session-prompt-generator
-description: Generating a ready-to-paste prompt for the next session — context, role, tasks, inputs, expected outputs
----
-# Session Prompt Generator — Generating a Prompt for the Next Session
-
-## When to Use
-- At the **end of a session** in Full Pipeline A (`/ship-right-thing`) or Full Pipeline B (`/shape-prioritize`) — before an inter-session `$handoff`.
-- When **handing the project to another operator** — forming a self-contained prompt.
-- When the **user requests** «generate a prompt to continue» — at any point.
-- During **emergency session termination** — preserving maximum context for recovery.
-
-> **Key principle:** the next session starts from **ZERO** context. The prompt must contain everything necessary for a new session to continue without loss.
-
-## Input
-
-| Field | Required | Description |
-|-------|:--------:|-------------|
-| `$board` state | ✅ | Current task board with all statuses |
-| Completed deliverables | ✅ | Complete artifacts from current and previous sessions |
-| Current gate | ✅ | Where we stopped (gate ID from `$board`) |
-| Next gate | ✅ | Where we are going (next gate ID) |
-| Adversarial camp | ⬚ | For Full A: Customer-Champion (α) / Business-Champion (β). For Full B: Build-Camp (α) / Cut-Camp (β). Determines artifact filtering |
-| Open questions | ⬚ | Blockers, unresolved conflicts, questions for the user |
-| Current session number | ✅ | Which session is ending (1-6) |
-| Initiative name | ✅ | Name of the product initiative |
-
-### Related Skills
-| Skill | Interaction | When |
-|-------|-------------|------|
-| `$board` | Source: current board state is included in the prompt (block 5) | Step 0 |
-| `$handoff` | Prompt is included in the inter-session handoff file | Step 5 |
-| `$gates` | Source: next gate requirements define the task (block 6) | Step 0 |
-| Framework skills | Completed frameworks ($prd-template, $rice-scoring etc.) included in full text in block 4 | Step 1 |
-
-## Session → Gates → Agents → Skills Mapping
-
-### Full Pipeline A (`/ship-right-thing`, Customer-vs-Business)
-
-| Session | Gates | Agents | Required Skills | Framework Skills |
-|:-------:|-------|--------|----------------|-----------------|
-| **1** | COND-01, DISC-01 | Conductor, Discovery | `$board`, `$gates`, `$handoff` | `$jtbd-canvas`, `$problem-statement`, `$assumption-mapping`, `$user-interview-script` |
-| **2** | COND-02, STRAT-01α | Conductor, Product Strategist (Customer-Champion) | `$board`, `$gates`, `$handoff` | `$product-vision`, `$okr-framework`, `$north-star-metric`, `$product-roadmap` |
-| **3** | COND-03, STRAT-02β | Conductor, Product Strategist (Business-Champion) | `$board`, `$gates`, `$handoff` | (same skills, business lens) |
-| **4** | COND-04, MED-01 | Conductor, Mediator | `$board`, `$gates`, `$handoff` | — |
-| **5** | COND-05, PM-SPEC, UX-01, TECH-01, DATA-01 | Conductor, PM, UX Designer, Tech Lead, Data Analyst | `$board`, `$gates`, `$handoff` | `$prd-template`, `$user-story`, `$acceptance-criteria`, `$user-flow`, `$design-brief`, `$epic-breakdown`, `$hypothesis-template`, `$saas-metrics`, `$ab-test-design` |
-| **6** | COND-06, DS-01, LY-01, RG-01 | Conductor, Designer, Layouter | `$board`, `$gates`, `$handoff`, `$report-design`, `$html-pdf-report`, `$launch-checklist` | — |
-
-### Full Pipeline B (`/shape-prioritize`, Build-vs-Cut)
-
-| Session | Gates | Agents | Required Skills | Framework Skills |
-|:-------:|-------|--------|----------------|-----------------|
-| **1** | COND-01, DISC-01 | Conductor, Discovery | `$board`, `$gates`, `$handoff` | `$problem-statement`, `$assumption-mapping` |
-| **2** | COND-02, PM-01α, TECH-01α | Conductor, PM (Build-Camp), Tech Lead (Build-Camp) | `$board`, `$gates`, `$handoff` | `$rice-scoring`, `$kano-model`, `$epic-breakdown` |
-| **3** | COND-03, PM-02β, TECH-02β | Conductor, PM (Cut-Camp), Tech Lead (Cut-Camp) | `$board`, `$gates`, `$handoff` | `$moscow-prioritization`, `$wsjf-scoring`, `$epic-breakdown` |
-| **4** | COND-04, MED-01 | Conductor, Mediator | `$board`, `$gates`, `$handoff` | — |
-| **5** | COND-05, PM-SPEC, UX-01, DATA-01 | Conductor, PM, UX Designer, Data Analyst | `$board`, `$gates`, `$handoff` | `$prd-template`, `$user-story`, `$acceptance-criteria`, `$saas-metrics` |
-| **6** | COND-06, DS-01, LY-01, RG-01 | Conductor, Designer, Layouter | `$board`, `$gates`, `$handoff`, `$report-design`, `$html-pdf-report` | — |
-
-### Quick Pipeline (`/quick-pm`) and Spec (`/spec`)
-Single session — prompt not needed (all gates within one session).
-
-## Protocol
-
-### Step 0 — State Collection
-1. Read the current `$board` state — all gates, statuses, deliverables.
-2. Collect all completed artifacts from the current session.
-3. Load the cumulative index from previous sessions (if available).
-4. Determine the next gate and its requirements from `$gates`.
-5. For Full A/B — determine the next session's camp (α / β / Mediator-neutral).
-6. Using the «Session → Gates → Agents → Skills» mapping determine what to load.
-
-### Step 1 — Identifying Critical Data
-1. Identify data that is **unrecoverable** without previous context:
-
-   | Category | Examples | Priority |
-   |----------|---------|:--------:|
-   | User responses | Discovery interviews (DISC-01), clarifications, decisions | 🔴 Critical |
-   | JTBD canvas / problems | Canvas, top 3 problem statements | 🔴 Critical |
-   | Evidence quotes | Verbatim customer quotes | 🔴 Critical |
-   | Strategy Briefs (Full A) | Alpha + Beta strategy | 🔴 Critical |
-   | Scope proposals (Full B) | Build + Cut proposals | 🔴 Critical |
-   | Mediator synthesis | Unified strategy / scope + rationale | 🔴 Critical |
-   | PRD content | User stories + AC | 🔴 Critical |
-   | User decisions | Approved, rejections, scope changes | 🟡 Important |
-   | Progress metrics | Time on gates, blockers | 🟢 Nice-to-have |
-
-2. **Rule:** all 🔴 Critical data is included in **full text**. No exceptions.
-3. **Adversarial filter (Full A/B):** if the next session is the opposing camp (Beta after Alpha, before Mediator), opposing camp artifacts are **EXCLUDED** from the prompt. This is a direct consequence of the camp independence rule.
-
-### Step 2 — Identifying Agents and Skills
-1. Using the mapping determine: which agents and skills are needed in the next session.
-2. Form a list of files to load (specific paths).
-3. For framework skills: specify the concrete skills if already known (from Strategy Brief or PRD).
-
-### Step 3 — Forming the Prompt
-1. Fill in the prompt template (6 blocks — see below).
-2. Include FULL TEXT of critical data (not links, not summaries).
-3. Formulate a specific task for the next session.
-4. Specify which agents and skills to activate (specific file paths).
-5. For Full A/B — explicitly state the camp (Alpha / Beta / neutral).
-
-### Step 4 — Prompt Validation
-Self-sufficiency test — for each point answer «can this be done without external context?»:
-
-| # | Check | Result |
-|---|-------|:------:|
-| 1 | Can work begin having ONLY this prompt? | Yes / No |
-| 2 | No references to «previous context» / «as discussed»? | Yes / No |
-| 3 | All 🔴 Critical data included in full text? | Yes / No |
-| 4 | Session task is specific (gate, action, deliverable)? | Yes / No |
-| 5 | Agents and skills listed with file paths? | Yes / No |
-| 6 | Task board included? | Yes / No |
-| 7 | Open questions recorded? | Yes / No |
-| 8 | Camp explicitly stated (Full A/B)? | Yes / No |
-| 9 | Opposing camp artifacts excluded (Full A/B)? | Yes / No |
-
-**If even one = No — revise the prompt.**
-
-### Step 5 — Saving
-1. Include the prompt in the `$handoff` file: `docs/product/session-N-handoff.md`.
-2. Show the prompt to the user for saving / copying.
-3. Warn: «This prompt — paste it at the start of the next session».
-
-### Step 6 — Emergency Mode
-If the session terminates unexpectedly (timeout, error, context loss):
-
-1. **Immediately** save everything available: board, last deliverable, open questions.
-2. Form an **emergency prompt** (minimum version):
-   - Block 1 (identification) — required.
-   - Block 5 (board state) — required.
-   - Block 4 (data) — everything that could be collected.
-   - Block 6 (task) — «Restore context from session-N-handoff.md, verify data integrity, continue from gate [ID]».
-3. Mark the prompt as **⚠️ EMERGENCY — possible data loss**.
-4. Indicate which data might have been lost and how to recover (repeat user interview, request from user).
-
-## Generated Prompt Structure
-
-### Block 1 — Initiative Identification
-- Initiative name.
-- Pipeline mode (/ship-right-thing / /shape-prioritize).
-- Next session number.
-- Which sessions are complete.
-- Next session camp (Alpha / Beta / Mediator-neutral / —).
-
-### Block 2 — Loading Instructions
-- Read `domains/product/AGENTS.md`.
-- Load agents needed in this session (from mapping).
-- Load skills: `$board`, `$handoff`, `$gates` + gate-specific ones.
-- Specific file paths.
-
-### Block 3 — Work Completed (summary)
-- What was done in previous sessions (2-3 sentences per session).
-- Key decisions and their rationale.
-
-### Block 4 — Critical Data
-- FULL TEXT of key artifacts needed to continue.
-- Discovery Brief, JTBD canvas, Strategy Briefs, Mediator synthesis.
-- User responses to interviews (if any).
-- **Camp filter:** opposing camp artifacts excluded (Full A/B, before Mediator).
-
-### Block 5 — Current State
-- Task board (`$board`) in current state.
-- Which gate is next.
-- Open questions and blockers.
-
-### Block 6 — Session Task
-- Specific formulation: what needs to be done in this session.
-- Expected outcome (deliverables).
-- Constraints and specifics (camp lens for Full A/B).
-
-## Prompt Formation Rules
-
-### Required Rules
-1. **Self-sufficiency** — the prompt must work without ANY previous context.
-2. **Data completeness** — include FULL text, not links and not summaries.
-3. **Task specificity** — not «continue work», but «execute STRAT-02β: form Business-Champion Strategy Brief based on Discovery Brief, ignore Camp Alpha artifacts».
-4. **Structure** — clear blocks, marked with headings.
-5. **Skills loading** — explicitly state which agents and skills to load with file paths.
-6. **Adversarial independence** — for Full A/B when handing off to opposing camp: explicitly record artifact exclusion.
-
-### Prohibited Elements
-1. References to «previous session» without full content.
-2. «As we discussed» — no context, nothing was discussed.
-3. «See results above» — there is no «above», everything must be in the prompt.
-4. Abbreviated data — «key conclusions: ...». ALL conclusions are needed.
-5. Assumptions about context — «you already know that...».
-6. In Full A/B: mixing artifacts from different camps before Mediator.
-
-## Example — Prompt for Session 3 (TeamFlow, `/ship-right-thing`, Beta Camp: Business-Champion)
-
-**Context:** TeamFlow (B2B SaaS HR-tech, $8M ARR). Initiative: AI-powered 1:1 summarization. Session 2 complete — Camp Alpha (Customer-Champion) formed their Strategy Brief. Session 3: Camp Beta (Business-Champion). KEY: Alpha artifacts are EXCLUDED.
-
-```
----
-## Prompt for Session 3 — AI-powered 1:1 Summarization (TeamFlow)
-
-### 1. Initiative
-- **Name:** AI-powered 1:1 note summarization for TeamFlow
-- **Mode:** Full Pipeline A (/ship-right-thing)
-- **Session:** 3 of 6
-- **Camp:** **Beta — Business-Champion**
-- **Previous sessions:** 1 (Discovery Brief), 2 (Strategy Brief Alpha: Customer-Champion, EXCLUDED from this prompt)
-
-### 2. Loading
-Read:
-- `domains/product/AGENTS.md`
-- `domains/product/.agents/skills/board/SKILL.md`
-- `domains/product/.agents/skills/handoff/SKILL.md`
-- `domains/product/.agents/skills/gates/SKILL.md`
-- `domains/product/.agents/skills/product-vision/SKILL.md`
-- `domains/product/.agents/skills/okr-framework/SKILL.md`
-- `domains/product/.agents/skills/north-star-metric/SKILL.md`
-- `domains/product/.agents/skills/product-roadmap/SKILL.md`
-- `domains/product/agents/product_strategist.md`
-- `domains/product/agents/conductor.md`
-
-Activate: Conductor (COND-03), Product Strategist (STRAT-02β).
-Role: **Business-Champion (β)** — focus on revenue, margin, competitive moat, strategic positioning. Do NOT play Customer-Champion.
-
-### 3. What Was Done
-**Session 1:** Discovery. Conducted 8 customer interviews (4 buyer VP HR + 4 end-user managers). JTBD canvas (buyer + end-user separately). Top 3 problem statements. Assumption map (12 assumptions, 4 high-risk).
-**Session 2:** Camp Alpha (Customer-Champion) formed their Strategy Brief. **IMPORTANT:** this brief is NOT passed to you until MED-01 (Step 4). You work independently, based on Discovery Brief ONLY.
-
-**Key decisions:**
-| # | Decision | Rationale | Who decided |
-|---|---------|-----------|-------------|
-| 1 | Focus on mid-market + enterprise (not SMB) | SMB pays less + higher churn | User, Session 1 |
-| 2 | Adversarial: Customer vs Business | High-stakes initiative, two perspectives insufficient | User, Session 1 |
-
-### 4. Data (full text, with camp filtering)
-
-#### Discovery Brief (from DISC-01 / session-1-handoff.md)
-
-**ICP:**
-- Firmographic: B2B SaaS / Tech companies, 100-500 employees, $5M-$50M ARR
-- Buyer persona: VP People/HR, 5-10 years experience, reports to CEO/COO
-- End-user persona: People managers (5-15 direct reports), some ICs in structured programs
-- Geographic: US (60%) + EU (30%) + APAC (10%)
-
-**JTBD (buyer = VP HR):**
-- **Functional:** «Ensure every manager runs consistently high-quality 1:1s across org»
-- **Emotional:** «Feel confident that people development does not depend on individual manager skill»
-- **Social:** «Be seen as data-driven People leader by CEO»
-
-**JTBD (end-user = manager):**
-- **Functional:** «Capture 1:1 discussion and not lose action items between meetings»
-- **Emotional:** «Feel prepared without spending 30 min before each 1:1»
-- **Social:** «Be seen as supportive manager by reports»
-
-**Top 3 Problem Statements:**
-1. Managers with 5+ reports spend 45-60 min/week on prep + note-taking, 30% of them acknowledge burnout from admin overhead. Evidence: 6 of 8 interviews + 120 support tickets Q1 2026.
-2. VP HR has no visibility into 1:1 quality/frequency across org — relying on self-report + anecdote. Evidence: 3 of 4 buyer interviews.
-3. Action items from 1:1s are not completed in 40% of cases (by self-report), often forgotten. Evidence: 5 of 8 end-user interviews + internal product analytics (action items marked completed vs total created).
-
-**Top 4 High-Risk Assumptions:**
-- **Value (V1):** Customers will pay $5-15/seat/month premium for AI summarization feature. Uncertainty 4, Impact 5, Risk 20.
-- **Feasibility (F1):** LLM summarization quality acceptable (no hallucinations) for sensitive HR conversations. Uncertainty 4, Impact 5, Risk 20.
-- **Viability (Vi1):** Privacy / GDPR / SOC2 compliance implications manageable for HR data ingestion. Uncertainty 3, Impact 5, Risk 15.
-- **Usability (U1):** Managers will trust AI summaries enough to use them instead of manual notes. Uncertainty 4, Impact 4, Risk 16.
-
-**Evidence Inventory:** 8 interviews (verbatim quotes attached in session-1-handoff.md), 120 support tickets Q1 2026, internal product analytics, competitor feature comparison (Lattice, 15Five, Culture Amp).
-
-**Open Questions:**
-- Pricing model: per-seat premium vs bundled vs usage-based?
-- Build vs Buy for LLM (OpenAI vs fine-tuned open-source)?
-- Scope: just 1:1s or all meetings (team meetings, reviews)?
-
-**🚫 Camp Alpha Strategy Brief — EXCLUDED from prompt (independence rule).**
-
-### 5. Current State
-
-#### Task Board
-📊 Progress: 29% (5/17) | Session: 3/6 | Blockers: 0
-
-| # | ID | Gate | Session | Camp | Status | Deliverable |
-|---|-----|------|:-------:|:----:|:------:|-------------|
-| 1 | COND-01 | Conductor | 1 | — | [✓] | Mode /ship-right-thing |
-| 2 | DISC-01 | Discovery | 1 | — | [✓] | Discovery Brief |
-| 3 | COND-02 | Conductor (s.2) | 2 | — | [✓] | Camp Alpha assigned |
-| 4 | STRAT-01α | Customer-Champion | 2 | α | [✓] | Strategy Brief Alpha (EXCLUDED) |
-| 5 | COND-03 | Conductor (s.3) | 3 | — | [→] | — |
-| 6 | STRAT-02β | Business-Champion | 3 | β | [ ] | — |
-| 7-17 | ... | ... | 4-6 | ... | [ ] | — |
-
-#### Open Questions
-- Pricing model (buyer interview gave no clear signal)
-- LLM build vs buy
-
-### 6. Session 3 Task (Beta Camp)
-
-**Gates:** COND-03 → STRAT-02β
-
-**What to do (Business-Champion lens):**
-1. **COND-03 (Conductor):** Load Discovery Brief, assign Camp Beta role explicitly. Verify Camp Alpha artifacts EXCLUDED (independence check).
-2. **STRAT-02β (Product Strategist):** Form Business-Champion Strategy Brief:
-   - **Vision** — business outcome frame («help companies measurably improve manager effectiveness and retain talent»)
-   - **NSM** — business-value metric (example: «ARR from accounts with AI summarization enabled on >60% teams»)
-   - **OKR** — quarterly goals with revenue / NRR / competitive moat emphasis
-   - **Roadmap** — themes with business ROI rationale
-   - **Pricing hypothesis** — specific model (per-seat premium? usage-based?)
-
-**Expected outcome:**
-`strategy-brief-beta.md` (~800-1200 words) with explicit business-lens argument. Include:
-- Why now (market timing)
-- Revenue impact model (3-year forecast scenarios)
-- Competitive moat angle
-- NRR / expansion story
-
-**Constraints:**
-- **DO NOT** pretend to play Customer-Champion lens
-- **DO NOT** request Camp Alpha artifacts (they're held by Conductor until MED-01)
-- Mediator (MED-01, Session 4) synthesizes — your task is to present the sharpest business case
-
-→ Start as Conductor (COND-03). Verify camp isolation, activate Product Strategist (β). Pass Discovery Brief (ONLY) as input.
----
-```
-
-## Validation (Quality Gate)
-
-A prompt is considered correct if:
-
-- [ ] Prompt is self-contained — work can begin without external context
-- [ ] Initiative name and mode stated
-- [ ] Session number and session history stated
-- [ ] Camp explicitly stated (Full A/B)
-- [ ] Agents and skills for loading listed with file paths (from mapping)
-- [ ] All 🔴 Critical data included in full text
-- [ ] Task board (`$board`) included in current state with metrics
-- [ ] Next gate and task formulated specifically (gate, action, deliverable)
-- [ ] Camp filter applied: opposing camp artifacts explicitly excluded (Full A/B, before Mediator)
-- [ ] Open questions and blockers recorded
-- [ ] No prohibited elements (references without content, abbreviations, assumptions)
-- [ ] Validation test (9 points, step 4) — all «Yes»
-- [ ] In emergency mode — prompt marked ⚠️, lost data indicated
-
-## Handoff
-The result of `$session-prompt-generator` is input for:
-- **Next session** — via user copying the prompt.
-- **`$handoff`** — prompt included in `docs/product/session-N-handoff.md` file.
-- **`$board`** — board state saved in the prompt.
-
-Transfer format: prompt in markdown, ready to paste at the start of a new session.
-
-## Anti-patterns
-
-| Error | Why it's bad | How to do it right |
-|-------|-------------|-------------------|
-| Prompt without Discovery data | Next session loses customer-centric foundation | Include FULL text of Discovery Brief |
-| «Continue work» | No context, nothing to continue from | Specific task: gate, action, deliverable |
-| File references without content | Files may be unavailable | Include content directly in the prompt |
-| Prompt without task board | Unclear where we are in the pipeline | Always include `$board` with metrics |
-| Abbreviated artifacts | Loss of critical details | Full text of each 🔴 Critical artifact |
-| Skipping open questions | Problems will resurface | Record all blockers and questions |
-| No loading instructions | Session won't load needed agents | Mapping → specific file paths |
-| Prompt > 50% «filler» | Dilutes critical data | Structure: blocks, tables, specifics |
-| Emergency termination without prompt | Total context loss | Emergency mode: minimal prompt + ⚠️ mark |
-| Camp contamination in prompt | Opposing camp artifacts enter prompt → independence violated | Explicit filter + note «EXCLUDED» for missing artifacts |
-
-## Output Template
-
-```
----
-## Prompt for Session [N+1] — [Initiative Name]
-
-### 1. Initiative
-- **Name:** [Name]
-- **Mode:** [/ship-right-thing | /shape-prioritize]
-- **Session:** [N+1] of 6
-- **Camp:** [Alpha (Customer/Build) | Beta (Business/Cut) | Mediator-neutral | —]
-- **Previous sessions:** 1-[N] complete
-
-### 2. Loading
-Read:
-- `domains/product/AGENTS.md`
-- `domains/product/.agents/skills/board/SKILL.md`
-- `domains/product/.agents/skills/handoff/SKILL.md`
-- `domains/product/.agents/skills/gates/SKILL.md`
-- [Specific skills: file paths]
-
-Activate agents: [List of agents with gate IDs]
-
-Role: **[Camp]** — [description of lens: focus on X, do not do Y]
-
-### 3. What Was Done
-**Session 1:** [Summary 2-3 sentences]
-**Session [N]:** [Summary 2-3 sentences]
-
-**Key decisions:**
-| # | Decision | Rationale | Who decided |
-|---|---------|-----------|-------------|
-| 1 | [Decision] | [Why] | [User / Conductor] |
-
-### 4. Data (full text, with camp filtering)
-
-#### Discovery Brief (from DISC-01)
-[FULL TEXT]
-
-#### Strategy Briefs (if Mediator Session)
-#### Alpha: [FULL TEXT]
-#### Beta: [FULL TEXT]
-
-#### [Other artifacts as needed]
-[FULL TEXT]
-
-**🚫 Excluded (camp filter):**
-- [Opposing camp artifacts, if applicable]
-
-### 5. Current State
-
-#### Task Board
-[Full board from $board with progress metrics]
-
-#### Open Questions
-| # | Question / Blocker | Severity | For which gate |
-|---|-------------------|:--------:|:--------------:|
-| 1 | [Description] | 🔴/🟡/🟢 | [ID] |
-
-### 6. Session [N+1] Task
-**Gates:** [ID list]
-**What to do:** [Specific description per gate + camp lens]
-**Expected outcome:** [Deliverables]
-**Constraints:** [Camp independence rules, DO NOT do, DO NOT request]
-
-→ Start as Conductor (COND-xx). Load board, verify camp isolation, pass control to [next agent].
----
-```
-
-### Emergency Prompt (minimum version)
-
-```
----
-## ⚠️ EMERGENCY PROMPT — Session [N+1] — [Initiative Name]
-
-**Status:** Session [N] terminated abnormally. Data loss possible.
-
-### What Was Saved
-[Everything that could be collected: board, last deliverable, questions]
-
-### What May Have Been Lost
-| # | Data | Recovery Method |
-|---|------|-----------------|
-| 1 | [What was lost] | [Repeat interview / request from user / restore from handoff] |
-
-### Task
-Restore context from `docs/product/session-[N]-handoff.md`. Verify data integrity. For Full A/B — verify camp isolation. Continue from gate [ID].
-
-→ Start as Conductor (COND-xx). Priority: data integrity verification before continuing.
----
-```
+---
+name: session-prompt-generator
+description: Generating a ready-to-paste prompt for the next session — context, role, tasks, inputs, expected outputs
+type: triggered
+domain: product
+owners:
+  - conductor
+gates:
+  - DISCOVERY
+  - PRODUCT_STRATEGIST
+  - PM
+  - UX_DESIGNER
+  - TECH_LEAD
+  - DATA_ANALYST
+  - MEDIATOR
+  - DESIGNER
+  - LAYOUTER
+  - RELEASE_GATE
+tech: []
+topic: []
+triggers:
+  - session-prompt-generator
+  - session prompt
+  - next session prompt
+related: []
+budget_lines: 474
+schema_version: 1
+---
+# Session Prompt Generator — Generating a Prompt for the Next Session
+
+## When to Use
+- At the **end of a session** in Full Pipeline A (`/ship-right-thing`) or Full Pipeline B (`/shape-prioritize`) — before an inter-session `$handoff`.
+- When **handing the project to another operator** — forming a self-contained prompt.
+- When the **user requests** «generate a prompt to continue» — at any point.
+- During **emergency session termination** — preserving maximum context for recovery.
+
+> **Key principle:** the next session starts from **ZERO** context. The prompt must contain everything necessary for a new session to continue without loss.
+
+## Input
+
+| Field | Required | Description |
+|-------|:--------:|-------------|
+| `$board` state | ✅ | Current task board with all statuses |
+| Completed deliverables | ✅ | Complete artifacts from current and previous sessions |
+| Current gate | ✅ | Where we stopped (gate ID from `$board`) |
+| Next gate | ✅ | Where we are going (next gate ID) |
+| Adversarial camp | ⬚ | For Full A: Customer-Champion (α) / Business-Champion (β). For Full B: Build-Camp (α) / Cut-Camp (β). Determines artifact filtering |
+| Open questions | ⬚ | Blockers, unresolved conflicts, questions for the user |
+| Current session number | ✅ | Which session is ending (1-6) |
+| Initiative name | ✅ | Name of the product initiative |
+
+### Related Skills
+| Skill | Interaction | When |
+|-------|-------------|------|
+| `$board` | Source: current board state is included in the prompt (block 5) | Step 0 |
+| `$handoff` | Prompt is included in the inter-session handoff file | Step 5 |
+| `$gates` | Source: next gate requirements define the task (block 6) | Step 0 |
+| Framework skills | Completed frameworks ($prd-template, $rice-scoring etc.) included in full text in block 4 | Step 1 |
+
+## Session → Gates → Agents → Skills Mapping
+
+### Full Pipeline A (`/ship-right-thing`, Customer-vs-Business)
+
+| Session | Gates | Agents | Required Skills | Framework Skills |
+|:-------:|-------|--------|----------------|-----------------|
+| **1** | COND-01, DISC-01 | Conductor, Discovery | `$board`, `$gates`, `$handoff` | `$jtbd-canvas`, `$problem-statement`, `$assumption-mapping`, `$user-interview-script` |
+| **2** | COND-02, STRAT-01α | Conductor, Product Strategist (Customer-Champion) | `$board`, `$gates`, `$handoff` | `$product-vision`, `$okr-framework`, `$north-star-metric`, `$product-roadmap` |
+| **3** | COND-03, STRAT-02β | Conductor, Product Strategist (Business-Champion) | `$board`, `$gates`, `$handoff` | (same skills, business lens) |
+| **4** | COND-04, MED-01 | Conductor, Mediator | `$board`, `$gates`, `$handoff` | — |
+| **5** | COND-05, PM-SPEC, UX-01, TECH-01, DATA-01 | Conductor, PM, UX Designer, Tech Lead, Data Analyst | `$board`, `$gates`, `$handoff` | `$prd-template`, `$user-story`, `$acceptance-criteria`, `$user-flow`, `$design-brief`, `$epic-breakdown`, `$hypothesis-template`, `$saas-metrics`, `$ab-test-design` |
+| **6** | COND-06, DS-01, LY-01, RG-01 | Conductor, Designer, Layouter | `$board`, `$gates`, `$handoff`, `$report-design`, `$html-pdf-report`, `$launch-checklist` | — |
+
+### Full Pipeline B (`/shape-prioritize`, Build-vs-Cut)
+
+| Session | Gates | Agents | Required Skills | Framework Skills |
+|:-------:|-------|--------|----------------|-----------------|
+| **1** | COND-01, DISC-01 | Conductor, Discovery | `$board`, `$gates`, `$handoff` | `$problem-statement`, `$assumption-mapping` |
+| **2** | COND-02, PM-01α, TECH-01α | Conductor, PM (Build-Camp), Tech Lead (Build-Camp) | `$board`, `$gates`, `$handoff` | `$rice-scoring`, `$kano-model`, `$epic-breakdown` |
+| **3** | COND-03, PM-02β, TECH-02β | Conductor, PM (Cut-Camp), Tech Lead (Cut-Camp) | `$board`, `$gates`, `$handoff` | `$moscow-prioritization`, `$wsjf-scoring`, `$epic-breakdown` |
+| **4** | COND-04, MED-01 | Conductor, Mediator | `$board`, `$gates`, `$handoff` | — |
+| **5** | COND-05, PM-SPEC, UX-01, DATA-01 | Conductor, PM, UX Designer, Data Analyst | `$board`, `$gates`, `$handoff` | `$prd-template`, `$user-story`, `$acceptance-criteria`, `$saas-metrics` |
+| **6** | COND-06, DS-01, LY-01, RG-01 | Conductor, Designer, Layouter | `$board`, `$gates`, `$handoff`, `$report-design`, `$html-pdf-report` | — |
+
+### Quick Pipeline (`/quick-pm`) and Spec (`/spec`)
+Single session — prompt not needed (all gates within one session).
+
+## Protocol
+
+### Step 0 — State Collection
+1. Read the current `$board` state — all gates, statuses, deliverables.
+2. Collect all completed artifacts from the current session.
+3. Load the cumulative index from previous sessions (if available).
+4. Determine the next gate and its requirements from `$gates`.
+5. For Full A/B — determine the next session's camp (α / β / Mediator-neutral).
+6. Using the «Session → Gates → Agents → Skills» mapping determine what to load.
+
+### Step 1 — Identifying Critical Data
+1. Identify data that is **unrecoverable** without previous context:
+
+   | Category | Examples | Priority |
+   |----------|---------|:--------:|
+   | User responses | Discovery interviews (DISC-01), clarifications, decisions | 🔴 Critical |
+   | JTBD canvas / problems | Canvas, top 3 problem statements | 🔴 Critical |
+   | Evidence quotes | Verbatim customer quotes | 🔴 Critical |
+   | Strategy Briefs (Full A) | Alpha + Beta strategy | 🔴 Critical |
+   | Scope proposals (Full B) | Build + Cut proposals | 🔴 Critical |
+   | Mediator synthesis | Unified strategy / scope + rationale | 🔴 Critical |
+   | PRD content | User stories + AC | 🔴 Critical |
+   | User decisions | Approved, rejections, scope changes | 🟡 Important |
+   | Progress metrics | Time on gates, blockers | 🟢 Nice-to-have |
+
+2. **Rule:** all 🔴 Critical data is included in **full text**. No exceptions.
+3. **Adversarial filter (Full A/B):** if the next session is the opposing camp (Beta after Alpha, before Mediator), opposing camp artifacts are **EXCLUDED** from the prompt. This is a direct consequence of the camp independence rule.
+
+### Step 2 — Identifying Agents and Skills
+1. Using the mapping determine: which agents and skills are needed in the next session.
+2. Form a list of files to load (specific paths).
+3. For framework skills: specify the concrete skills if already known (from Strategy Brief or PRD).
+
+### Step 3 — Forming the Prompt
+1. Fill in the prompt template (6 blocks — see below).
+2. Include FULL TEXT of critical data (not links, not summaries).
+3. Formulate a specific task for the next session.
+4. Specify which agents and skills to activate (specific file paths).
+5. For Full A/B — explicitly state the camp (Alpha / Beta / neutral).
+
+### Step 4 — Prompt Validation
+Self-sufficiency test — for each point answer «can this be done without external context?»:
+
+| # | Check | Result |
+|---|-------|:------:|
+| 1 | Can work begin having ONLY this prompt? | Yes / No |
+| 2 | No references to «previous context» / «as discussed»? | Yes / No |
+| 3 | All 🔴 Critical data included in full text? | Yes / No |
+| 4 | Session task is specific (gate, action, deliverable)? | Yes / No |
+| 5 | Agents and skills listed with file paths? | Yes / No |
+| 6 | Task board included? | Yes / No |
+| 7 | Open questions recorded? | Yes / No |
+| 8 | Camp explicitly stated (Full A/B)? | Yes / No |
+| 9 | Opposing camp artifacts excluded (Full A/B)? | Yes / No |
+
+**If even one = No — revise the prompt.**
+
+### Step 5 — Saving
+1. Include the prompt in the `$handoff` file: `docs/product/session-N-handoff.md`.
+2. Show the prompt to the user for saving / copying.
+3. Warn: «This prompt — paste it at the start of the next session».
+
+### Step 6 — Emergency Mode
+If the session terminates unexpectedly (timeout, error, context loss):
+
+1. **Immediately** save everything available: board, last deliverable, open questions.
+2. Form an **emergency prompt** (minimum version):
+   - Block 1 (identification) — required.
+   - Block 5 (board state) — required.
+   - Block 4 (data) — everything that could be collected.
+   - Block 6 (task) — «Restore context from session-N-handoff.md, verify data integrity, continue from gate [ID]».
+3. Mark the prompt as **⚠️ EMERGENCY — possible data loss**.
+4. Indicate which data might have been lost and how to recover (repeat user interview, request from user).
+
+## Generated Prompt Structure
+
+### Block 1 — Initiative Identification
+- Initiative name.
+- Pipeline mode (/ship-right-thing / /shape-prioritize).
+- Next session number.
+- Which sessions are complete.
+- Next session camp (Alpha / Beta / Mediator-neutral / —).
+
+### Block 2 — Loading Instructions
+- Read `domains/product/AGENTS.md`.
+- Load agents needed in this session (from mapping).
+- Load skills: `$board`, `$handoff`, `$gates` + gate-specific ones.
+- Specific file paths.
+
+### Block 3 — Work Completed (summary)
+- What was done in previous sessions (2-3 sentences per session).
+- Key decisions and their rationale.
+
+### Block 4 — Critical Data
+- FULL TEXT of key artifacts needed to continue.
+- Discovery Brief, JTBD canvas, Strategy Briefs, Mediator synthesis.
+- User responses to interviews (if any).
+- **Camp filter:** opposing camp artifacts excluded (Full A/B, before Mediator).
+
+### Block 5 — Current State
+- Task board (`$board`) in current state.
+- Which gate is next.
+- Open questions and blockers.
+
+### Block 6 — Session Task
+- Specific formulation: what needs to be done in this session.
+- Expected outcome (deliverables).
+- Constraints and specifics (camp lens for Full A/B).
+
+## Prompt Formation Rules
+
+### Required Rules
+1. **Self-sufficiency** — the prompt must work without ANY previous context.
+2. **Data completeness** — include FULL text, not links and not summaries.
+3. **Task specificity** — not «continue work», but «execute STRAT-02β: form Business-Champion Strategy Brief based on Discovery Brief, ignore Camp Alpha artifacts».
+4. **Structure** — clear blocks, marked with headings.
+5. **Skills loading** — explicitly state which agents and skills to load with file paths.
+6. **Adversarial independence** — for Full A/B when handing off to opposing camp: explicitly record artifact exclusion.
+
+### Prohibited Elements
+1. References to «previous session» without full content.
+2. «As we discussed» — no context, nothing was discussed.
+3. «See results above» — there is no «above», everything must be in the prompt.
+4. Abbreviated data — «key conclusions: ...». ALL conclusions are needed.
+5. Assumptions about context — «you already know that...».
+6. In Full A/B: mixing artifacts from different camps before Mediator.
+
+## Example — Prompt for Session 3 (TeamFlow, `/ship-right-thing`, Beta Camp: Business-Champion)
+
+**Context:** TeamFlow (B2B SaaS HR-tech, $8M ARR). Initiative: AI-powered 1:1 summarization. Session 2 complete — Camp Alpha (Customer-Champion) formed their Strategy Brief. Session 3: Camp Beta (Business-Champion). KEY: Alpha artifacts are EXCLUDED.
+
+```
+---
+## Prompt for Session 3 — AI-powered 1:1 Summarization (TeamFlow)
+
+### 1. Initiative
+- **Name:** AI-powered 1:1 note summarization for TeamFlow
+- **Mode:** Full Pipeline A (/ship-right-thing)
+- **Session:** 3 of 6
+- **Camp:** **Beta — Business-Champion**
+- **Previous sessions:** 1 (Discovery Brief), 2 (Strategy Brief Alpha: Customer-Champion, EXCLUDED from this prompt)
+
+### 2. Loading
+Read:
+- `domains/product/AGENTS.md`
+- `domains/product/.agents/skills/board/SKILL.md`
+- `domains/product/.agents/skills/handoff/SKILL.md`
+- `domains/product/.agents/skills/gates/SKILL.md`
+- `domains/product/.agents/skills/product-vision/SKILL.md`
+- `domains/product/.agents/skills/okr-framework/SKILL.md`
+- `domains/product/.agents/skills/north-star-metric/SKILL.md`
+- `domains/product/.agents/skills/product-roadmap/SKILL.md`
+- `domains/product/agents/product_strategist.md`
+- `domains/product/agents/conductor.md`
+
+Activate: Conductor (COND-03), Product Strategist (STRAT-02β).
+Role: **Business-Champion (β)** — focus on revenue, margin, competitive moat, strategic positioning. Do NOT play Customer-Champion.
+
+### 3. What Was Done
+**Session 1:** Discovery. Conducted 8 customer interviews (4 buyer VP HR + 4 end-user managers). JTBD canvas (buyer + end-user separately). Top 3 problem statements. Assumption map (12 assumptions, 4 high-risk).
+**Session 2:** Camp Alpha (Customer-Champion) formed their Strategy Brief. **IMPORTANT:** this brief is NOT passed to you until MED-01 (Step 4). You work independently, based on Discovery Brief ONLY.
+
+**Key decisions:**
+| # | Decision | Rationale | Who decided |
+|---|---------|-----------|-------------|
+| 1 | Focus on mid-market + enterprise (not SMB) | SMB pays less + higher churn | User, Session 1 |
+| 2 | Adversarial: Customer vs Business | High-stakes initiative, two perspectives insufficient | User, Session 1 |
+
+### 4. Data (full text, with camp filtering)
+
+#### Discovery Brief (from DISC-01 / session-1-handoff.md)
+
+**ICP:**
+- Firmographic: B2B SaaS / Tech companies, 100-500 employees, $5M-$50M ARR
+- Buyer persona: VP People/HR, 5-10 years experience, reports to CEO/COO
+- End-user persona: People managers (5-15 direct reports), some ICs in structured programs
+- Geographic: US (60%) + EU (30%) + APAC (10%)
+
+**JTBD (buyer = VP HR):**
+- **Functional:** «Ensure every manager runs consistently high-quality 1:1s across org»
+- **Emotional:** «Feel confident that people development does not depend on individual manager skill»
+- **Social:** «Be seen as data-driven People leader by CEO»
+
+**JTBD (end-user = manager):**
+- **Functional:** «Capture 1:1 discussion and not lose action items between meetings»
+- **Emotional:** «Feel prepared without spending 30 min before each 1:1»
+- **Social:** «Be seen as supportive manager by reports»
+
+**Top 3 Problem Statements:**
+1. Managers with 5+ reports spend 45-60 min/week on prep + note-taking, 30% of them acknowledge burnout from admin overhead. Evidence: 6 of 8 interviews + 120 support tickets Q1 2026.
+2. VP HR has no visibility into 1:1 quality/frequency across org — relying on self-report + anecdote. Evidence: 3 of 4 buyer interviews.
+3. Action items from 1:1s are not completed in 40% of cases (by self-report), often forgotten. Evidence: 5 of 8 end-user interviews + internal product analytics (action items marked completed vs total created).
+
+**Top 4 High-Risk Assumptions:**
+- **Value (V1):** Customers will pay $5-15/seat/month premium for AI summarization feature. Uncertainty 4, Impact 5, Risk 20.
+- **Feasibility (F1):** LLM summarization quality acceptable (no hallucinations) for sensitive HR conversations. Uncertainty 4, Impact 5, Risk 20.
+- **Viability (Vi1):** Privacy / GDPR / SOC2 compliance implications manageable for HR data ingestion. Uncertainty 3, Impact 5, Risk 15.
+- **Usability (U1):** Managers will trust AI summaries enough to use them instead of manual notes. Uncertainty 4, Impact 4, Risk 16.
+
+**Evidence Inventory:** 8 interviews (verbatim quotes attached in session-1-handoff.md), 120 support tickets Q1 2026, internal product analytics, competitor feature comparison (Lattice, 15Five, Culture Amp).
+
+**Open Questions:**
+- Pricing model: per-seat premium vs bundled vs usage-based?
+- Build vs Buy for LLM (OpenAI vs fine-tuned open-source)?
+- Scope: just 1:1s or all meetings (team meetings, reviews)?
+
+**🚫 Camp Alpha Strategy Brief — EXCLUDED from prompt (independence rule).**
+
+### 5. Current State
+
+#### Task Board
+📊 Progress: 29% (5/17) | Session: 3/6 | Blockers: 0
+
+| # | ID | Gate | Session | Camp | Status | Deliverable |
+|---|-----|------|:-------:|:----:|:------:|-------------|
+| 1 | COND-01 | Conductor | 1 | — | [✓] | Mode /ship-right-thing |
+| 2 | DISC-01 | Discovery | 1 | — | [✓] | Discovery Brief |
+| 3 | COND-02 | Conductor (s.2) | 2 | — | [✓] | Camp Alpha assigned |
+| 4 | STRAT-01α | Customer-Champion | 2 | α | [✓] | Strategy Brief Alpha (EXCLUDED) |
+| 5 | COND-03 | Conductor (s.3) | 3 | — | [→] | — |
+| 6 | STRAT-02β | Business-Champion | 3 | β | [ ] | — |
+| 7-17 | ... | ... | 4-6 | ... | [ ] | — |
+
+#### Open Questions
+- Pricing model (buyer interview gave no clear signal)
+- LLM build vs buy
+
+### 6. Session 3 Task (Beta Camp)
+
+**Gates:** COND-03 → STRAT-02β
+
+**What to do (Business-Champion lens):**
+1. **COND-03 (Conductor):** Load Discovery Brief, assign Camp Beta role explicitly. Verify Camp Alpha artifacts EXCLUDED (independence check).
+2. **STRAT-02β (Product Strategist):** Form Business-Champion Strategy Brief:
+   - **Vision** — business outcome frame («help companies measurably improve manager effectiveness and retain talent»)
+   - **NSM** — business-value metric (example: «ARR from accounts with AI summarization enabled on >60% teams»)
+   - **OKR** — quarterly goals with revenue / NRR / competitive moat emphasis
+   - **Roadmap** — themes with business ROI rationale
+   - **Pricing hypothesis** — specific model (per-seat premium? usage-based?)
+
+**Expected outcome:**
+`strategy-brief-beta.md` (~800-1200 words) with explicit business-lens argument. Include:
+- Why now (market timing)
+- Revenue impact model (3-year forecast scenarios)
+- Competitive moat angle
+- NRR / expansion story
+
+**Constraints:**
+- **DO NOT** pretend to play Customer-Champion lens
+- **DO NOT** request Camp Alpha artifacts (they're held by Conductor until MED-01)
+- Mediator (MED-01, Session 4) synthesizes — your task is to present the sharpest business case
+
+→ Start as Conductor (COND-03). Verify camp isolation, activate Product Strategist (β). Pass Discovery Brief (ONLY) as input.
+---
+```
+
+## Validation (Quality Gate)
+
+A prompt is considered correct if:
+
+- [ ] Prompt is self-contained — work can begin without external context
+- [ ] Initiative name and mode stated
+- [ ] Session number and session history stated
+- [ ] Camp explicitly stated (Full A/B)
+- [ ] Agents and skills for loading listed with file paths (from mapping)
+- [ ] All 🔴 Critical data included in full text
+- [ ] Task board (`$board`) included in current state with metrics
+- [ ] Next gate and task formulated specifically (gate, action, deliverable)
+- [ ] Camp filter applied: opposing camp artifacts explicitly excluded (Full A/B, before Mediator)
+- [ ] Open questions and blockers recorded
+- [ ] No prohibited elements (references without content, abbreviations, assumptions)
+- [ ] Validation test (9 points, step 4) — all «Yes»
+- [ ] In emergency mode — prompt marked ⚠️, lost data indicated
+
+## Handoff
+The result of `$session-prompt-generator` is input for:
+- **Next session** — via user copying the prompt.
+- **`$handoff`** — prompt included in `docs/product/session-N-handoff.md` file.
+- **`$board`** — board state saved in the prompt.
+
+Transfer format: prompt in markdown, ready to paste at the start of a new session.
+
+## Anti-patterns
+
+| Error | Why it's bad | How to do it right |
+|-------|-------------|-------------------|
+| Prompt without Discovery data | Next session loses customer-centric foundation | Include FULL text of Discovery Brief |
+| «Continue work» | No context, nothing to continue from | Specific task: gate, action, deliverable |
+| File references without content | Files may be unavailable | Include content directly in the prompt |
+| Prompt without task board | Unclear where we are in the pipeline | Always include `$board` with metrics |
+| Abbreviated artifacts | Loss of critical details | Full text of each 🔴 Critical artifact |
+| Skipping open questions | Problems will resurface | Record all blockers and questions |
+| No loading instructions | Session won't load needed agents | Mapping → specific file paths |
+| Prompt > 50% «filler» | Dilutes critical data | Structure: blocks, tables, specifics |
+| Emergency termination without prompt | Total context loss | Emergency mode: minimal prompt + ⚠️ mark |
+| Camp contamination in prompt | Opposing camp artifacts enter prompt → independence violated | Explicit filter + note «EXCLUDED» for missing artifacts |
+
+## Output Template
+
+```
+---
+## Prompt for Session [N+1] — [Initiative Name]
+
+### 1. Initiative
+- **Name:** [Name]
+- **Mode:** [/ship-right-thing | /shape-prioritize]
+- **Session:** [N+1] of 6
+- **Camp:** [Alpha (Customer/Build) | Beta (Business/Cut) | Mediator-neutral | —]
+- **Previous sessions:** 1-[N] complete
+
+### 2. Loading
+Read:
+- `domains/product/AGENTS.md`
+- `domains/product/.agents/skills/board/SKILL.md`
+- `domains/product/.agents/skills/handoff/SKILL.md`
+- `domains/product/.agents/skills/gates/SKILL.md`
+- [Specific skills: file paths]
+
+Activate agents: [List of agents with gate IDs]
+
+Role: **[Camp]** — [description of lens: focus on X, do not do Y]
+
+### 3. What Was Done
+**Session 1:** [Summary 2-3 sentences]
+**Session [N]:** [Summary 2-3 sentences]
+
+**Key decisions:**
+| # | Decision | Rationale | Who decided |
+|---|---------|-----------|-------------|
+| 1 | [Decision] | [Why] | [User / Conductor] |
+
+### 4. Data (full text, with camp filtering)
+
+#### Discovery Brief (from DISC-01)
+[FULL TEXT]
+
+#### Strategy Briefs (if Mediator Session)
+#### Alpha: [FULL TEXT]
+#### Beta: [FULL TEXT]
+
+#### [Other artifacts as needed]
+[FULL TEXT]
+
+**🚫 Excluded (camp filter):**
+- [Opposing camp artifacts, if applicable]
+
+### 5. Current State
+
+#### Task Board
+[Full board from $board with progress metrics]
+
+#### Open Questions
+| # | Question / Blocker | Severity | For which gate |
+|---|-------------------|:--------:|:--------------:|
+| 1 | [Description] | 🔴/🟡/🟢 | [ID] |
+
+### 6. Session [N+1] Task
+**Gates:** [ID list]
+**What to do:** [Specific description per gate + camp lens]
+**Expected outcome:** [Deliverables]
+**Constraints:** [Camp independence rules, DO NOT do, DO NOT request]
+
+→ Start as Conductor (COND-xx). Load board, verify camp isolation, pass control to [next agent].
+---
+```
+
+### Emergency Prompt (minimum version)
+
+```
+---
+## ⚠️ EMERGENCY PROMPT — Session [N+1] — [Initiative Name]
+
+**Status:** Session [N] terminated abnormally. Data loss possible.
+
+### What Was Saved
+[Everything that could be collected: board, last deliverable, questions]
+
+### What May Have Been Lost
+| # | Data | Recovery Method |
+|---|------|-----------------|
+| 1 | [What was lost] | [Repeat interview / request from user / restore from handoff] |
+
+### Task
+Restore context from `docs/product/session-[N]-handoff.md`. Verify data integrity. For Full A/B — verify camp isolation. Continue from gate [ID].
+
+→ Start as Conductor (COND-xx). Priority: data integrity verification before continuing.
+---
+```