code-ai-installer 4.0.0 → 4.0.1-b

package/domains/product/locales/en/.agents/skills/product-roadmap/SKILL.md

@@ -1,305 +1,321 @@
----
-name: product-roadmap
-description: Theme-based roadmap — Now / Next / Later, without hard dates, attached to OKR
----
-# Product Roadmap
-
-> **Category:** Strategy  ·  **Slug:** `product-roadmap`
-
-## When to Use
-
-- During quarterly or semi-annual planning.
-- For stakeholder communication (exec, sales, customer success).
-- For alignment of cross-functional work (eng + design + marketing + CS).
-- For a public-facing customer roadmap (with caveats).
-
-## Input
-
-| Field | Required | Description |
-|-------|:--------:|-------------|
-| Product vision | ✅ | Long-term direction |
-| OKRs | ✅ | Current quarter goals |
-| North Star Metric | ✅ | Alignment anchor |
-| Backlog (prioritized) | ✅ | Initiatives + score |
-| Capacity | ✅ | Team size, eng bandwidth |
-| Known commitments | ⬚ | Customer promises, compliance deadlines |
-
-## Data Sources
-
-1. `$okr-framework` — quarterly goals.
-2. `$rice-scoring` / `$moscow-prioritization` — initiatives.
-3. `$jtbd-canvas` / `$opportunity-solution-tree` — opportunity themes.
-4. Customer feedback (support, NPS, direct requests).
-
-### Related Skills
-
-| Skill | What we take | When to invoke |
-|-------|-------------|----------------|
-| `okr-framework` | KRs | Before roadmap |
-| `rice-scoring` | Prioritized initiatives | For Now/Next/Later assignment |
-| `kano-model` | Feature balance | Balance must-have / delighters |
-| `product-vision` | Themes rationale | Themes must fit vision |
-
-## Format: Now / Next / Later by Theme
-
-```
-Theme A: [Customer outcome]
-  Now (0-3 months):    [Initiative 1], [Initiative 2]
-  Next (3-6 months):   [Initiative 3]
-  Later (6+ months):   [Theme direction]
-
-Theme B: ...
-```
-
-**Themes**, not features: each theme = customer outcome or strategic direction.
-
-## Protocol
-
-### Step 0 — Identify Themes
-
-From vision + OKR determine 3-5 themes for the current horizon.
-
-Theme = customer-valued outcome, not «Q2 release».
-
-✅ Good themes:
-- «Onboarding: new teams activate in <7 days»
-- «Enterprise-ready: SSO, SCIM, audit logs»
-- «Usage visibility: account-level analytics»
-
-❌ Bad themes:
-- «Platform improvements» (too abstract)
-- «Bug fixes» (not a theme, this is operational)
-- «Ship feature X» (single feature ≠ theme)
-
-### Step 1 — Initiatives per Theme
-
-For each theme — a list of initiatives (from backlog). Via `$rice-scoring`.
-
-| Theme | Initiative | RICE Score | Effort (t-shirt) |
-|-------|------------|:----------:|:----------------:|
-| Onboarding | In-app checklist | 45 | M |
-| Onboarding | Team templates | 30 | L |
-
-### Step 2 — Capacity Check
-
-- Team size × weeks = total person-weeks
-- Minus: oncall, meetings, eng debt, slack (~30%)
-- Net available = useful capacity
-
-Match Now (0-3 months) to ~70-80% capacity. The rest — buffer for incidents, emergent work.
-
-### Step 3 — Assignment to Now / Next / Later
-
-**Now** — committed to work, specific scope, start date within 6 weeks.
-
-**Next** — high-priority, but scope/timing TBD. Will start in 3-6 months.
-
-**Later** — strategically important, but not yet shaped. Placeholder for directions.
-
-Don't put dates, except for hard external commits (compliance, customer SLA).
-
-### Step 4 — Cross-team Dependencies
-
-For each initiative:
-- Which teams are involved (design, data, platform, security)?
-- External dependencies (vendors, legal, partnerships)?
-- Hard deadlines (contract, regulation)?
-
-### Step 5 — Risks + Contingencies
-
-- Top 3 risks per theme (missing deps, scope uncertainty, hiring)
-- Trigger events — what will force a roadmap adjustment
-
-### Step 6 — Communication Artifact
-
-Two formats:
-- **Internal roadmap** — full, with RICE scores, risks
-- **External roadmap** (if shared with customers) — themes + directions, **without promised dates**, with disclaimer «subject to change»
-
-### Step 7 — Review Cadence
-
-- **Monthly:** progress, emergent work, re-assignment
-- **Quarterly:** Now → shipped, Next → Now, Later → shaped
-- **Board/exec:** quarterly summary
-
-## Validation (Quality Gate)
-
-- [ ] 3-5 themes, customer-outcome framed
-- [ ] Each theme has initiatives
-- [ ] Initiatives — scored (RICE / MoSCoW)
-- [ ] Capacity check: Now ~70-80%
-- [ ] No hard dates except external commits
-- [ ] Dependencies identified
-- [ ] Risks listed
-- [ ] Internal + external versions separated
-
-## Handoff
-
-The result is input for:
-- **PM** → PRDs for Now initiatives
-- **Tech Lead** → engineering planning
-- **Stakeholders** → expectation setting
-- **Mediator** (Full A) — for synthesis
-
-Format: roadmap doc (markdown + optional visual / timeline). Via `$handoff`.
-
-## Anti-patterns
-
-| Error | Why it's bad | How to do it right |
-|-------|-------------|-------------------|
-| Dated roadmap (public) | Missed dates → trust erosion | Themes without dates externally |
-| Feature list instead of themes | No outcome framing | Themes = outcomes |
-| 100% capacity | No buffer for incidents | 70-80% in Now |
-| Too many themes (6+) | Focus lost | 3-5 themes |
-| No review cadence | Stale roadmap | Monthly check-ins |
-| Promise «all of this» | Under-delivers | Now = commitment, Next = intent, Later = direction |
-
-## Template
-
-```markdown
-# Product Roadmap — Q2-Q3 2026
-
-## Themes
-
-### Theme 1: [Customer Outcome]
-**Why:** [alignment with vision / OKR]
-
-#### Now (0-3 months)
-- [Initiative A] — RICE 45, Effort M, Owner: PM-X
-- [Initiative B] — RICE 30, Effort L
-
-#### Next (3-6 months)
-- [Initiative C]
-
-#### Later (6+ months)
-- [Direction]
-
-**Dependencies:** [cross-team]
-**Risks:** [top 3]
-
-### Theme 2: ...
-
----
-
-## Capacity
-- Team: N people
-- Now commitment: X person-weeks (Y% capacity)
-- Buffer: 20%
-
-## Review Cadence
-- Monthly: progress + emergent work
-- Quarterly: re-shape Next/Later
-```
-
-## Worked Example — TeamFlow Roadmap Q2-Q3 2026
-
-**Context:** TeamFlow after Q1 Mediator synthesis + OKR set. Roadmap aligns to unified vision (workflow + analytics), cascades through 3 Objectives.
-
-```markdown
-# Product Roadmap — Q2-Q3 2026 (TeamFlow Product)
-
-## Strategic Context
-- **Vision:** AI-powered People Ops platform (workflow + analytics)
-- **NSM:** Sticky Manager Ratio (AI Tier)
-- **OKRs:** O1 (AI tier ships + adoption), O2 (Enterprise tier expansion), O3 (data foundation)
-- **Horizon:** 6 months (Q2 + Q3)
-
-## Themes
-
-### Theme 1: AI-Assisted 1:1 Workflow (End-User Value)
-**Why:** Drives O1 (manager adoption). Addresses top JTBD (manager admin overhead).
-
-#### Now (0-3 months, April-June)
-- **AI Summarization MVP** — auto-transcribe + summarize + action items. RICE 288, L effort. Owner: PM Alex. *Status: Week 4/12.*
-- **Inline Summary Edit** — manager reviews/corrects before save. Must-have from Kano. RICE 52, M effort.
-- **Action Items Auto-Extract + Reminders** — extracted from AI summary, surfaced before next 1:1. RICE 58, M.
-- **Per-Meeting AI Toggle** — privacy Must-have. RICE 45, S.
-
-#### Next (3-6 months, July-September)
-- **Real-Time Prep Card** — AI generates pre-1:1 context brief (previous topics, open action items, suggested prompts). Delighter-turning-Performance.
-- **Calendar Integration v2** — auto-pull from Google Cal + Outlook, map to TeamFlow 1:1s.
-- **Mobile Summary Review** — review / approve summaries from mobile.
-
-#### Later (6+ months, Q4+)
-- **Voice-only interaction** — manager speaks, no typing required (delighter).
-- **Multi-language summarization** — non-English companies.
-- **Group meeting summarization** — beyond 1:1s (scope creep controlled).
-
-**Dependencies:** LLM API budget approval (CFO, by Week 3); Privacy/Legal review for EU data residency.
-**Risks:** (1) LLM quality plateau below 85% acceptable — mitigation: human-review layer feature flag. (2) OpenAI pricing change — mitigation: vendor alternative tested (Anthropic via AWS Bedrock).
-
----
-
-### Theme 2: People Ops Analytics (Buyer Value → Enterprise Expansion)
-**Why:** Drives O2 (enterprise tier). Addresses VP HR visibility gap.
-
-#### Now (0-3 months)
-- **Aggregate 1:1 Cadence Dashboard** — per-team, per-manager frequency + duration. RICE 45, M.
-- **Manager Health Score (MVP)** — 4-component score (cadence, action items velocity, NPS, attrition). Delighter-turning-Performance. RICE 32, L.
-- **Quarterly Exec Report Generator** — auto-composed board-ready PDF. RICE 28, M.
-
-#### Next (3-6 months)
-- **Benchmark Engine** — industry comparisons (is our 1:1 cadence top-quartile?). Differentiation moat.
-- **At-Risk Team Detector** — pattern-based early warning. AI-powered.
-- **Manager Coaching Insights** — AI surfaces coaching opportunities per manager.
-
-#### Later (6+ months)
-- **Predictive Attrition Model** — manager practice data → retention risk.
-- **Cross-Customer Benchmarks** — network effect analytics.
-
-**Dependencies:** Data pipeline instrumentation (Theme 3 prerequisite). Security attestation SOC 2 Type II for analytics tier.
-**Risks:** (1) Benchmark data thin until N ≥ 20 Enterprise customers — mitigation: start with internal team benchmarks. (2) Manager backlash if perceived as surveillance — mitigation: aggregate-only, transparent data model.
-
----
-
-### Theme 3: Data & Privacy Foundation
-**Why:** Drives O3 (foundation). Enables Themes 1 & 2.
-
-#### Now (0-3 months)
-- **Event Schema + Instrumentation** — reliable 1:1 behavioral data pipeline. Platform work.
-- **Privacy Architecture** — opt-in, user-controlled, regional LLM endpoints (US + EU + APAC).
-- **SOC 2 Type II Extension** — to cover AI features.
-
-#### Next (3-6 months)
-- **Data Retention Controls UI** — customer-controlled (30/90/365 days).
-- **Audit Log Enhancements** — Who accessed what summary when.
-
-#### Later (6+ months)
-- **HIPAA-ready tier** — for healthcare customers.
-- **On-premise LLM option** — for ultra-regulated enterprise.
-
-**Dependencies:** Legal review; external security audit firm engagement.
-**Risks:** (1) Audit delays → Enterprise deals slip — mitigation: parallel-track audit + conditional deals.
-
----
-
-## Capacity
-
-- **Team:** 8 engineers + 3 PM + 2 design + 1 data = 14 people
-- **Available person-weeks:** 14 × 13 weeks = 182 weeks (per 13-sprint quarter)
-- **Allocations:**
-  - Theme 1: 70 person-weeks (54% of Now)
-  - Theme 2: 45 person-weeks (35% of Now)
-  - Theme 3: 30 person-weeks (23% — platform, invisible but critical)
-  - Total Now: 145 person-weeks
-  - **Buffer:** 37 person-weeks (20% — for incidents, emergent work, recruiting ramp)
-
-## Public-facing Roadmap (Customer-visible)
-
-### What's Next on TeamFlow
-
-- **Q2 2026:** AI-assisted 1:1 summaries + action items tracking (Team tier)
-- **Q3 2026:** People Ops analytics + team benchmarks (Enterprise tier)
-- **Coming Later:** Advanced AI coaching insights, predictive analytics
-
-*This public roadmap shows themes without specific dates. Features ship when ready; we share 
-progress through quarterly customer webinars. [Subscribe for updates].*
-
-## Review Cadence
-- **Monthly 1st Tuesday:** cross-theme progress + emergent work integration
-- **Mid-Q (Week 6, May 12):** re-calibrate Now, convert Next→Now if ahead
-- **End-Q (June 30):** full retro, Next→Now promotions, Later additions
-- **Annual (December):** Themes evolved / retired; new horizon defined
-```
-
-> **Roadmap lesson:** **Three themes = three OKRs**. Each theme drives one Objective, so the team can say «if I touch Theme 1 this week, I'm moving KR1.x». No orphan work. Dependencies between themes (Theme 3 enables Theme 2's data layer) are documented explicitly — this prevents the «why is the platform team busy?» confusion. The public-facing version strictly avoids dates — learned from prior missed-date backlash in Q4 2025.
+---
+name: product-roadmap
+description: Theme-based roadmap — Now / Next / Later, without hard dates, attached to OKR
+type: triggered
+domain: product
+owners:
+  - pm
+gates:
+  - PM
+tech: []
+topic: []
+triggers:
+  - product-roadmap
+  - roadmap
+  - дорожная карта
+  - now next later
+related: []
+budget_lines: 321
+schema_version: 1
+---
+# Product Roadmap
+
+> **Category:** Strategy  ·  **Slug:** `product-roadmap`
+
+## When to Use
+
+- During quarterly or semi-annual planning.
+- For stakeholder communication (exec, sales, customer success).
+- For alignment of cross-functional work (eng + design + marketing + CS).
+- For a public-facing customer roadmap (with caveats).
+
+## Input
+
+| Field | Required | Description |
+|-------|:--------:|-------------|
+| Product vision | ✅ | Long-term direction |
+| OKRs | ✅ | Current quarter goals |
+| North Star Metric | ✅ | Alignment anchor |
+| Backlog (prioritized) | ✅ | Initiatives + score |
+| Capacity | ✅ | Team size, eng bandwidth |
+| Known commitments | ⬚ | Customer promises, compliance deadlines |
+
+## Data Sources
+
+1. `$okr-framework` — quarterly goals.
+2. `$rice-scoring` / `$moscow-prioritization` — initiatives.
+3. `$jtbd-canvas` / `$opportunity-solution-tree` — opportunity themes.
+4. Customer feedback (support, NPS, direct requests).
+
+### Related Skills
+
+| Skill | What we take | When to invoke |
+|-------|-------------|----------------|
+| `okr-framework` | KRs | Before roadmap |
+| `rice-scoring` | Prioritized initiatives | For Now/Next/Later assignment |
+| `kano-model` | Feature balance | Balance must-have / delighters |
+| `product-vision` | Themes rationale | Themes must fit vision |
+
+## Format: Now / Next / Later by Theme
+
+```
+Theme A: [Customer outcome]
+  Now (0-3 months):    [Initiative 1], [Initiative 2]
+  Next (3-6 months):   [Initiative 3]
+  Later (6+ months):   [Theme direction]
+
+Theme B: ...
+```
+
+**Themes**, not features: each theme = customer outcome or strategic direction.
+
+## Protocol
+
+### Step 0 — Identify Themes
+
+From vision + OKR determine 3-5 themes for the current horizon.
+
+Theme = customer-valued outcome, not «Q2 release».
+
+✅ Good themes:
+- «Onboarding: new teams activate in <7 days»
+- «Enterprise-ready: SSO, SCIM, audit logs»
+- «Usage visibility: account-level analytics»
+
+❌ Bad themes:
+- «Platform improvements» (too abstract)
+- «Bug fixes» (not a theme, this is operational)
+- «Ship feature X» (single feature ≠ theme)
+
+### Step 1 — Initiatives per Theme
+
+For each theme — a list of initiatives (from backlog). Via `$rice-scoring`.
+
+| Theme | Initiative | RICE Score | Effort (t-shirt) |
+|-------|------------|:----------:|:----------------:|
+| Onboarding | In-app checklist | 45 | M |
+| Onboarding | Team templates | 30 | L |
+
+### Step 2 — Capacity Check
+
+- Team size × weeks = total person-weeks
+- Minus: oncall, meetings, eng debt, slack (~30%)
+- Net available = useful capacity
+
+Match Now (0-3 months) to ~70-80% capacity. The rest — buffer for incidents, emergent work.
+
+### Step 3 — Assignment to Now / Next / Later
+
+**Now** — committed to work, specific scope, start date within 6 weeks.
+
+**Next** — high-priority, but scope/timing TBD. Will start in 3-6 months.
+
+**Later** — strategically important, but not yet shaped. Placeholder for directions.
+
+Don't put dates, except for hard external commits (compliance, customer SLA).
+
+### Step 4 — Cross-team Dependencies
+
+For each initiative:
+- Which teams are involved (design, data, platform, security)?
+- External dependencies (vendors, legal, partnerships)?
+- Hard deadlines (contract, regulation)?
+
+### Step 5 — Risks + Contingencies
+
+- Top 3 risks per theme (missing deps, scope uncertainty, hiring)
+- Trigger events — what will force a roadmap adjustment
+
+### Step 6 — Communication Artifact
+
+Two formats:
+- **Internal roadmap** — full, with RICE scores, risks
+- **External roadmap** (if shared with customers) — themes + directions, **without promised dates**, with disclaimer «subject to change»
+
+### Step 7 — Review Cadence
+
+- **Monthly:** progress, emergent work, re-assignment
+- **Quarterly:** Now → shipped, Next → Now, Later → shaped
+- **Board/exec:** quarterly summary
+
+## Validation (Quality Gate)
+
+- [ ] 3-5 themes, customer-outcome framed
+- [ ] Each theme has initiatives
+- [ ] Initiatives — scored (RICE / MoSCoW)
+- [ ] Capacity check: Now ~70-80%
+- [ ] No hard dates except external commits
+- [ ] Dependencies identified
+- [ ] Risks listed
+- [ ] Internal + external versions separated
+
+## Handoff
+
+The result is input for:
+- **PM** → PRDs for Now initiatives
+- **Tech Lead** → engineering planning
+- **Stakeholders** → expectation setting
+- **Mediator** (Full A) — for synthesis
+
+Format: roadmap doc (markdown + optional visual / timeline). Via `$handoff`.
+
+## Anti-patterns
+
+| Error | Why it's bad | How to do it right |
+|-------|-------------|-------------------|
+| Dated roadmap (public) | Missed dates → trust erosion | Themes without dates externally |
+| Feature list instead of themes | No outcome framing | Themes = outcomes |
+| 100% capacity | No buffer for incidents | 70-80% in Now |
+| Too many themes (6+) | Focus lost | 3-5 themes |
+| No review cadence | Stale roadmap | Monthly check-ins |
+| Promise «all of this» | Under-delivers | Now = commitment, Next = intent, Later = direction |
+
+## Template
+
+```markdown
+# Product Roadmap — Q2-Q3 2026
+
+## Themes
+
+### Theme 1: [Customer Outcome]
+**Why:** [alignment with vision / OKR]
+
+#### Now (0-3 months)
+- [Initiative A] — RICE 45, Effort M, Owner: PM-X
+- [Initiative B] — RICE 30, Effort L
+
+#### Next (3-6 months)
+- [Initiative C]
+
+#### Later (6+ months)
+- [Direction]
+
+**Dependencies:** [cross-team]
+**Risks:** [top 3]
+
+### Theme 2: ...
+
+---
+
+## Capacity
+- Team: N people
+- Now commitment: X person-weeks (Y% capacity)
+- Buffer: 20%
+
+## Review Cadence
+- Monthly: progress + emergent work
+- Quarterly: re-shape Next/Later
+```
+
+## Worked Example — TeamFlow Roadmap Q2-Q3 2026
+
+**Context:** TeamFlow after Q1 Mediator synthesis + OKR set. Roadmap aligns to unified vision (workflow + analytics), cascades through 3 Objectives.
+
+```markdown
+# Product Roadmap — Q2-Q3 2026 (TeamFlow Product)
+
+## Strategic Context
+- **Vision:** AI-powered People Ops platform (workflow + analytics)
+- **NSM:** Sticky Manager Ratio (AI Tier)
+- **OKRs:** O1 (AI tier ships + adoption), O2 (Enterprise tier expansion), O3 (data foundation)
+- **Horizon:** 6 months (Q2 + Q3)
+
+## Themes
+
+### Theme 1: AI-Assisted 1:1 Workflow (End-User Value)
+**Why:** Drives O1 (manager adoption). Addresses top JTBD (manager admin overhead).
+
+#### Now (0-3 months, April-June)
+- **AI Summarization MVP** — auto-transcribe + summarize + action items. RICE 288, L effort. Owner: PM Alex. *Status: Week 4/12.*
+- **Inline Summary Edit** — manager reviews/corrects before save. Must-have from Kano. RICE 52, M effort.
+- **Action Items Auto-Extract + Reminders** — extracted from AI summary, surfaced before next 1:1. RICE 58, M.
+- **Per-Meeting AI Toggle** — privacy Must-have. RICE 45, S.
+
+#### Next (3-6 months, July-September)
+- **Real-Time Prep Card** — AI generates pre-1:1 context brief (previous topics, open action items, suggested prompts). Delighter-turning-Performance.
+- **Calendar Integration v2** — auto-pull from Google Cal + Outlook, map to TeamFlow 1:1s.
+- **Mobile Summary Review** — review / approve summaries from mobile.
+
+#### Later (6+ months, Q4+)
+- **Voice-only interaction** — manager speaks, no typing required (delighter).
+- **Multi-language summarization** — non-English companies.
+- **Group meeting summarization** — beyond 1:1s (scope creep controlled).
+
+**Dependencies:** LLM API budget approval (CFO, by Week 3); Privacy/Legal review for EU data residency.
+**Risks:** (1) LLM quality plateau below 85% acceptable — mitigation: human-review layer feature flag. (2) OpenAI pricing change — mitigation: vendor alternative tested (Anthropic via AWS Bedrock).
+
+---
+
+### Theme 2: People Ops Analytics (Buyer Value → Enterprise Expansion)
+**Why:** Drives O2 (enterprise tier). Addresses VP HR visibility gap.
+
+#### Now (0-3 months)
+- **Aggregate 1:1 Cadence Dashboard** — per-team, per-manager frequency + duration. RICE 45, M.
+- **Manager Health Score (MVP)** — 4-component score (cadence, action items velocity, NPS, attrition). Delighter-turning-Performance. RICE 32, L.
+- **Quarterly Exec Report Generator** — auto-composed board-ready PDF. RICE 28, M.
+
+#### Next (3-6 months)
+- **Benchmark Engine** — industry comparisons (is our 1:1 cadence top-quartile?). Differentiation moat.
+- **At-Risk Team Detector** — pattern-based early warning. AI-powered.
+- **Manager Coaching Insights** — AI surfaces coaching opportunities per manager.
+
+#### Later (6+ months)
+- **Predictive Attrition Model** — manager practice data → retention risk.
+- **Cross-Customer Benchmarks** — network effect analytics.
+
+**Dependencies:** Data pipeline instrumentation (Theme 3 prerequisite). Security attestation SOC 2 Type II for analytics tier.
+**Risks:** (1) Benchmark data thin until N ≥ 20 Enterprise customers — mitigation: start with internal team benchmarks. (2) Manager backlash if perceived as surveillance — mitigation: aggregate-only, transparent data model.
+
+---
+
+### Theme 3: Data & Privacy Foundation
+**Why:** Drives O3 (foundation). Enables Themes 1 & 2.
+
+#### Now (0-3 months)
+- **Event Schema + Instrumentation** — reliable 1:1 behavioral data pipeline. Platform work.
+- **Privacy Architecture** — opt-in, user-controlled, regional LLM endpoints (US + EU + APAC).
+- **SOC 2 Type II Extension** — to cover AI features.
+
+#### Next (3-6 months)
+- **Data Retention Controls UI** — customer-controlled (30/90/365 days).
+- **Audit Log Enhancements** — Who accessed what summary when.
+
+#### Later (6+ months)
+- **HIPAA-ready tier** — for healthcare customers.
+- **On-premise LLM option** — for ultra-regulated enterprise.
+
+**Dependencies:** Legal review; external security audit firm engagement.
+**Risks:** (1) Audit delays → Enterprise deals slip — mitigation: parallel-track audit + conditional deals.
+
+---
+
+## Capacity
+
+- **Team:** 8 engineers + 3 PM + 2 design + 1 data = 14 people
+- **Available person-weeks:** 14 × 13 weeks = 182 weeks (per 13-sprint quarter)
+- **Allocations:**
+  - Theme 1: 70 person-weeks (54% of Now)
+  - Theme 2: 45 person-weeks (35% of Now)
+  - Theme 3: 30 person-weeks (23% — platform, invisible but critical)
+  - Total Now: 145 person-weeks
+  - **Buffer:** 37 person-weeks (20% — for incidents, emergent work, recruiting ramp)
+
+## Public-facing Roadmap (Customer-visible)
+
+### What's Next on TeamFlow
+
+- **Q2 2026:** AI-assisted 1:1 summaries + action items tracking (Team tier)
+- **Q3 2026:** People Ops analytics + team benchmarks (Enterprise tier)
+- **Coming Later:** Advanced AI coaching insights, predictive analytics
+
+*This public roadmap shows themes without specific dates. Features ship when ready; we share 
+progress through quarterly customer webinars. [Subscribe for updates].*
+
+## Review Cadence
+- **Monthly 1st Tuesday:** cross-theme progress + emergent work integration
+- **Mid-Q (Week 6, May 12):** re-calibrate Now, convert Next→Now if ahead
+- **End-Q (June 30):** full retro, Next→Now promotions, Later additions
+- **Annual (December):** Themes evolved / retired; new horizon defined
+```
+
+> **Roadmap lesson:** **Three themes = three OKRs**. Each theme drives one Objective, so the team can say «if I touch Theme 1 this week, I'm moving KR1.x». No orphan work. Dependencies between themes (Theme 3 enables Theme 2's data layer) are documented explicitly — this prevents the «why is the platform team busy?» confusion. The public-facing version strictly avoids dates — learned from prior missed-date backlash in Q4 2025.