code-ai-installer 2.2.0 → 2.4.0

package/domains/product/locales/en/.agents/skills/north-star-metric/SKILL.md

@@ -0,0 +1,301 @@
+---
+name: north-star-metric
+description: Selecting the NSM for B2B SaaS — balancing leading/lagging, alignment with customer value and business outcome
+---
+# North Star Metric (NSM)
+
+> **Category:** Strategy  ·  **Slug:** `north-star-metric`
+
+## When to Use
+
+- When defining product strategy — a single metric of focus.
+- When aligning the team: decision-making through the lens of "does this lift the NSM?".
+- For board / investor reporting — single metric story.
+- When reviewing an existing metric — still meaningful?
+
+## Input
+
+| Field | Required | Description |
+|-------|:--------:|-------------|
+| Product vision | ✅ | The north star must reflect the vision |
+| Business model | ✅ | Subscription / usage / hybrid |
+| Current metric stack | ✅ | MRR, DAU, retention, etc. |
+| Customer value definition | ✅ | What we consider value for the customer |
+| Adversarial camp (Full A) | ⬚ | alpha (Customer-first) / beta (Business-first) |
+
+## Data Sources
+
+1. Current product analytics (Amplitude, Mixpanel, custom).
+2. Customer interviews → what "valuable" means.
+3. Business metrics — NRR, churn.
+4. Industry benchmarks (SaaS Capital, OpenView reports).
+
+### Related Skills
+
+| Skill | What we take | When to invoke |
+|-------|-------------|----------------|
+| `product-vision` | Vision → NSM alignment | Before choosing NSM |
+| `aarrr-metrics` | Framework for decomposition | For input metrics |
+| `saas-metrics` | Business metric layer | For lagging alignment |
+| `okr-framework` | OKR KRs → often tied to NSM | After NSM |
+
+## Criteria for a Good NSM
+
+1. **Reflects customer value** — when NSM grows, customers receive value.
+2. **Leading indicator of revenue** — moves before MRR.
+3. **Actionable** — the team can influence it.
+4. **Singular** — not composite, not aggregate.
+5. **Measurable** — numerically trackable.
+6. **Resistant to gaming** — hard to inflate without real value.
+
+## Protocol
+
+### Step 0 — Role Framing (Full A)
+
+- **Customer-Champion (alpha):** emphasis on user-value metric
+- **Business-Champion (beta):** emphasis on business-value metric
+
+### Step 1 — Business Model Analysis
+
+- **Subscription SaaS:** NSM often related to weekly/monthly active teams, usage depth
+- **Usage-based:** raw consumption (API calls, workloads)
+- **Transaction-based:** successful transactions
+- **Platform:** marketplace size × match rate
+
+### Step 2 — Generate Candidates
+
+Minimum 5 candidates. Each — through the formula:
+
+**[Action verb]** × **[Unit with value]** × **[Frequency/Cadence]** × **[Quality]**
+
+B2B SaaS examples:
+- "Weekly Active Teams with ≥5 seats" (team platform)
+- "Monthly reports generated per paying account" (analytics tool)
+- "Weekly successful deployments per engineering team" (DevOps tool)
+- "Onboarded workspaces completing setup within 7 days" (product-led growth)
+
+### Step 3 — Evaluation
+
+| Metric | Customer Value | Leads Revenue | Actionable | Singular | Measurable | Anti-game | Total |
+|--------|:--------------:|:-------------:|:----------:|:--------:|:----------:|:---------:|:-----:|
+| [cand] | 5 | 4 | 4 | 5 | 5 | 3 | 26 |
+
+Rank. Top candidate = NSM. Second → guardrail or input metric.
+
+### Step 4 — Decomposition (Metric Tree)
+
+NSM → input metrics → leading indicators.
+
+```
+NSM: Weekly Active Teams ≥5 seats
+  ├── Input: New teams activating weekly
+  │     ├── Signup → activation rate
+  │     └── Time-to-5-seats (onboarding speed)
+  ├── Input: Existing teams retaining
+  │     ├── W/W retention rate
+  │     └── Seats-per-team growth
+  └── Guardrails:
+        ├── Churn rate (capped)
+        ├── NPS (no drop)
+        └── Support ticket volume (no spike)
+```
+
+### Step 5 — Guardrails
+
+What **MUST NOT break** in the pursuit of NSM:
+- Churn rate
+- NPS / CSAT
+- Gross margin
+- Support ticket volume
+- p95 latency / error rate
+
+Each guardrail = numeric threshold.
+
+### Step 6 — Baseline + Target
+
+- Current value
+- 90-day target
+- 1-year aspirational
+
+### Step 7 — Anti-Gaming Review
+
+Team brainstorm: "How would we game the NSM?". If it's easy — enhance the definition or change the metric.
+
+Example: "Weekly Active Teams" can be gamed by counting churned teams that were active for the last time → clarification: "with activity in last 7 days".
+
+## Validation (Quality Gate)
+
+- [ ] Minimum 5 candidates reviewed
+- [ ] Evaluation across 6 criteria
+- [ ] Metric tree decomposition (NSM → inputs)
+- [ ] Guardrails with numeric thresholds
+- [ ] Baseline + target
+- [ ] Anti-gaming review conducted
+- [ ] Business model alignment verified
+- [ ] Vision alignment verified
+
+## Handoff
+
+The result is input for:
+- **`okr-framework`** — NSM = basis for several KRs
+- **`aarrr-metrics`** — NSM fits into AARRR as the key stage
+- **Data Analyst** — instrumentation requirements
+- **Mediator** (Full A) — alpha and beta NSM for synthesis
+
+Format: NSM doc (one-pager) with metric tree. Via `$handoff`.
+
+## Anti-patterns
+
+| Error | Why it's bad | How to do it right |
+|-------|-------------|-------------------|
+| Revenue as NSM | Lagging, not actionable | Use leading value metric |
+| Vanity metric | Looks good, doesn't predict | Customer value + revenue proxy |
+| Composite metric | Weighted sum is not singular | Pick ONE |
+| No guardrails | Accept churn growing with NSM | Explicit guardrails |
+| Change every quarter | Team can't align | Stable for ≥ year |
+| Copy other company | Context-specific | Your business model |
+
+## Template
+
+```markdown
+# North Star Metric: [Name]
+
+## Definition
+**[Action]** × **[Unit]** × **[Frequency]** × **[Quality]**
+Precise: [exact definition with inclusion/exclusion rules]
+
+## Rationale
+[Why this metric — customer value + business]
+
+## Metric Tree
+NSM
+├── Input: ...
+├── Input: ...
+└── Guardrails:
+    ├── ...
+
+## Baseline & Targets
+- Current: X
+- 90-day target: Y
+- 1-year aspirational: Z
+
+## Anti-Gaming
+[How could this be gamed? Mitigation]
+
+## Review
+- Quarterly: still meaningful?
+- Annually: re-evaluate
+```
+
+## Worked Example — TeamFlow NSM Selection (post-Mediator synthesis)
+
+**Context:** After Mediator synthesis of visions, the team chooses a unified NSM. Alpha proposed "weekly 1:1s with AI summaries used"; Beta — "ARR from accounts with ≥3 seats on AI tier". Below — full candidate evaluation.
+
+### Business Model Analysis
+
+- **Revenue model:** Subscription, tiered (Core / Team / Enterprise)
+- **Growth lever:** Seat expansion (mid-market, enterprise) + tier upgrade
+- **Primary value delivery:** Per-1:1 workflow savings + per-manager adoption
+
+### 5 Candidate NSMs
+
+| # | Candidate | Formula | Leading/Lagging | Ties to vision? |
+|---|-----------|---------|:---------------:|:---------------:|
+| C1 | Weekly Active Managers using AI summarization (≥ 1 per week) | count(WAM_ai) | Leading | Alpha ✓ |
+| C2 | ARR from AI-tier accounts | $ ARR tagged «AI tier» | Lagging | Beta ✓ |
+| C3 | «Sticky Manager Ratio» — managers running ≥ 4 AI-assisted 1:1s per month | ratio | Leading + quality signal | Alpha ✓ |
+| C4 | «AI-assisted 1:1s completed per week» (total count across base) | count | Leading, usage-based | Alpha ✓ |
+| C5 | Net New AI Tier ARR / Quarter | $ | Lagging | Beta ✓ |
+
+### Evaluation Matrix (6 criteria × 5 candidates, scale 1-5)
+
+| Metric | Customer Value | Leads Revenue | Actionable | Singular | Measurable | Anti-Game | **Total** |
+|--------|:--------------:|:-------------:|:----------:|:--------:|:----------:|:---------:|:---------:|
+| C1 WAM_ai | 4 | 4 | 5 | 5 | 5 | 3 | **26** |
+| C2 ARR AI tier | 2 | 2 (= it IS revenue) | 3 | 5 | 5 | 4 | 21 |
+| C3 Sticky Manager Ratio | 5 | 4 | 5 | 5 | 5 | 4 | **28** |
+| C4 AI 1:1s/week | 4 | 4 | 4 | 5 | 5 | 2 (easy to bulk) | 24 |
+| C5 Net New AI ARR | 2 | 2 (= it IS revenue) | 2 | 5 | 5 | 4 | 20 |
+
+**Winner: C3 «Sticky Manager Ratio»** — balances customer value signal (quality — 4+ 1:1s/month implies real habit) + anti-gaming (can't be bulk-created; requires genuine usage) + clear leading indicator for ARR.
+
+### Chosen NSM — Definition
+
+```markdown
+# North Star Metric: Sticky Manager Ratio (AI Tier)
+
+## Definition
+**Active managers** × **Using AI summarization** × **≥ 4 times per month** × **on ≥ 1 direct report**
+
+**Precise:**
+- **Active manager:** has ≥ 1 direct report in TeamFlow + logged in in last 30 days
+- **Using AI summarization:** completed 1:1 with AI summary generated AND reviewed (not just generated but ignored)
+- **≥ 4 per month:** rolling 30-day window
+- **Exclusions:** internal TeamFlow employees; accounts <30 days old (onboarding); trial accounts
+
+**Formula:**
+```
+Sticky Manager Ratio = Managers meeting criteria / Total active managers in AI-tier accounts
+```
+
+## Rationale
+
+**Why this metric:**
+
+- **Customer value signal:** 4 uses/month ≈ weekly 1:1s = genuine habit, not experimentation. Habit = product embedded in workflow → high retention.
+- **Leading revenue:** Sticky managers → account retention → expansion → NRR. Historical TeamFlow data: features used ≥ 4×/month have 3× LTV.
+- **Actionable:** Team can drive (onboarding, reminders, feature quality).
+- **Singular:** One number, clear dashboard.
+- **Resistant to gaming:** can't «bulk» real usage. Can be inflated via trial seats, so trial accounts excluded.
+
+**Why not alternatives:**
+- C1 WAM_ai: doesn't distinguish casual from habitual usage
+- C2 ARR tier: lagging, doesn't predict expansion before it happens
+- C4 1:1s/week: easily gamed («create 20 fake 1:1s»)
+- C5 Net new ARR: same lagging issue
+
+## Metric Tree
+
+```
+NSM: Sticky Manager Ratio (AI Tier)
+  ├── Input: AI Tier account count (expansion)
+  │     ├── New AI Tier conversions per week
+  │     └── AI Tier churn
+  ├── Input: Active manager activation rate (% new managers reaching ≥ 4 1:1s)
+  │     ├── Time to first AI summary
+  │     └── Onboarding checklist completion
+  ├── Input: Feature quality signal
+  │     ├── Summary approval rate (kept vs edited >50%)
+  │     └── Summary NPS (in-product rating)
+  └── Guardrails:
+        ├── Overall NPS (must not drop > 5 points)
+        ├── Gross churn rate (must not grow >1pp QoQ)
+        ├── Support ticket volume / AI issues (monitor for abuse / dissatisfaction)
+        └── p95 summary generation latency (< 30 sec)
+```
+
+## Baseline & Targets
+
+| Horizon | Target | Rationale |
+|---------|:------:|-----------|
+| Current (Q1 2026) | 0% (feature not live) | Pre-launch |
+| 90 days post-launch (end Q2) | **20%** | Conservative — based on similar AI feature launches in adjacent B2B |
+| End of 2026 | 45% | Category-leading level |
+| End of 2027 (aspirational) | 65% | «Cannot imagine working without it» level |
+
+## Anti-Gaming Analysis
+
+**How could this be gamed?**
+
+- **Bulk fake 1:1 creation:** Mitigation — require AI summary generated AND reviewed (visible edit or approval event)
+- **Single massive account inflating ratio:** Mitigation — weight by account size + report separately
+- **Managers «maintaining streak» with perfunctory 1:1s:** Mitigation — 1:1 duration floor (must be ≥ 10 min scheduled)
+- **Team inflating by marking non-actual 1:1s:** Mitigation — periodic sample audit with customer success conversations
+
+## Review
+- **Weekly:** ratio tracked in exec dashboard
+- **Quarterly:** anti-gaming sample audit (3 accounts reviewed manually)
+- **Annually:** NSM re-evaluated against strategy shift
+```
+
+> **NSM lesson:** Alpha's initial «WAM_ai» was good but too blunt — didn't distinguish trial adoption from habit. Beta's «ARR AI tier» was too lagging. Synthesis («Sticky Manager Ratio») borrowed Alpha's user-centric formulation but applied Beta's rigor about quality threshold (4+ per month = real habit). Result is a metric that **leads** revenue, **measures** customer value, and **resists** easy gaming. This is the kind of refinement that's hard to get on first try — the evaluation matrix forced the conversation.

package/domains/product/locales/en/.agents/skills/north-star-metric/agents/claude.json

@@ -0,0 +1,16 @@
+{
+  "name": "north-star-metric",
+  "display_name": "North Star Metric",
+  "description": "Selecting the NSM for B2B SaaS — balancing leading/lagging, alignment with customer value and business outcome",
+  "default_prompt": "Use $north-star-metric when the task relates to the skill \"North Star Metric\".",
+  "triggers": [
+    "north-star-metric",
+    "north star",
+    "NSM"
+  ],
+  "capabilities": [
+    "north-star-metric"
+  ],
+  "tools": [],
+  "implicit_invocation": true
+}

package/domains/product/locales/en/.agents/skills/north-star-metric/agents/copilot.json

@@ -0,0 +1,16 @@
+{
+  "name": "north-star-metric",
+  "display_name": "North Star Metric",
+  "description": "Selecting the NSM for B2B SaaS — balancing leading/lagging, alignment with customer value and business outcome",
+  "default_prompt": "Use $north-star-metric when the task relates to the skill \"North Star Metric\".",
+  "triggers": [
+    "north-star-metric",
+    "north star",
+    "NSM"
+  ],
+  "capabilities": [
+    "north-star-metric"
+  ],
+  "tools": [],
+  "implicit_invocation": true
+}

package/domains/product/locales/en/.agents/skills/north-star-metric/agents/gemini.json

@@ -0,0 +1,16 @@
+{
+  "name": "north-star-metric",
+  "display_name": "North Star Metric",
+  "description": "Selecting the NSM for B2B SaaS — balancing leading/lagging, alignment with customer value and business outcome",
+  "default_prompt": "Use $north-star-metric when the task relates to the skill \"North Star Metric\".",
+  "triggers": [
+    "north-star-metric",
+    "north star",
+    "NSM"
+  ],
+  "capabilities": [
+    "north-star-metric"
+  ],
+  "tools": [],
+  "implicit_invocation": true
+}

package/domains/product/locales/en/.agents/skills/north-star-metric/agents/openai.yaml

@@ -0,0 +1,8 @@
+interface:
+  display_name: "North Star Metric (NSM)"
+  short_description: "Selecting the NSM for B2B SaaS — balancing leading/lagging, alignment with customer value and business outcome"
+  default_prompt: "Use $north-star-metric when the task relates to the skill \"North Star Metric (NSM)\"."
+dependencies:
+  tools: []
+policy:
+  allow_implicit_invocation: true

package/domains/product/locales/en/.agents/skills/north-star-metric/agents/qwen.json

@@ -0,0 +1,16 @@
+{
+  "name": "north-star-metric",
+  "display_name": "North Star Metric",
+  "description": "Selecting the NSM for B2B SaaS — balancing leading/lagging, alignment with customer value and business outcome",
+  "default_prompt": "Use $north-star-metric when the task relates to the skill \"North Star Metric\".",
+  "triggers": [
+    "north-star-metric",
+    "north star",
+    "NSM"
+  ],
+  "capabilities": [
+    "north-star-metric"
+  ],
+  "tools": [],
+  "implicit_invocation": true
+}

package/domains/product/locales/en/.agents/skills/north-star-metric/agents/skill.yaml

@@ -0,0 +1,19 @@
+version: 1
+name: "north-star-metric"
+display_name: "North Star Metric"
+description: "Selecting the NSM for B2B SaaS — balancing leading/lagging, alignment with customer value and business outcome"
+default_prompt: "Use $north-star-metric when the task relates to the skill \"North Star Metric (NSM)\"."
+triggers:
+  - "north-star-metric"
+  - "north star"
+  - "NSM"
+capabilities:
+  - "north-star-metric"
+tools: []
+invocation:
+  explicit: true
+  implicit: true
+localization:
+  default_locale: "en"
+  available_locales:
+    - "en"

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

@@ -0,0 +1,284 @@
+---
+name: okr-framework
+description: Objectives & Key Results for B2B product teams — quarterly cycle, measurable key results
+---
+# OKR Framework
+
+> **Category:** Strategy  ·  **Slug:** `okr-framework`
+
+## When to Use
+
+- During quarterly planning for the product team.
+- When translating vision → operational goals.
+- When aligning with exec-level goals (cascading OKR).
+- During post-quarter review — grading KRs.
+
+## Input
+
+| Field | Required | Description |
+|-------|:--------:|-------------|
+| Product vision | ✅ | What is the long-term north star |
+| North Star Metric | ✅ | Measurable anchor |
+| Company-level goals | ✅ | Which OKRs come from above |
+| Previous quarter grade | ⬚ | For adjustments |
+| Team capacity | ✅ | Honest assessment of resources |
+
+## Data Sources
+
+1. `$product-vision` — long-term direction.
+2. `$north-star-metric` — primary metric.
+3. Company OKR — alignment with company goals.
+4. Previous quarter results — learning.
+
+### Related Skills
+
+| Skill | What we take | When to invoke |
+|-------|-------------|----------------|
+| `product-vision` | Vision | Before OKR — for alignment |
+| `north-star-metric` | NSM | KRs often tied to NSM |
+| `product-roadmap` | Initiatives | OKR → roadmap themes |
+| `rice-scoring` | Priorities | Which initiatives this quarter |
+
+## Structure
+
+```
+Objective (qualitative goal, inspirational)
+  ├── Key Result 1 (measurable, outcome, not output)
+  ├── Key Result 2
+  └── Key Result 3
+```
+
+## Protocol
+
+### Step 0 — Cascade from Company
+
+1. Review company-level Objectives.
+2. Which of them does our product team drive?
+3. Alignment is not through copy-paste — through "how we contribute".
+
+### Step 1 — Objectives (1-3)
+
+Rules:
+- **Qualitative, inspirational** — not a number
+- Clear enough to know «did we achieve it?»
+- Time-bounded (usually a quarter)
+- 1-3 per quarter — more = distraction
+
+✅ Good: «Become the default onboarding tool for mid-market engineering teams»  
+❌ Bad: «Increase feature adoption» (weakly inspirational) or «Grow ARR by 15%» (this is a KR, not an O)
+
+### Step 2 — Key Results per Objective (2-5)
+
+Rules:
+- **Measurable** — numeric target
+- **Outcome, not output** (not «ship feature X», but «feature X used by Y teams»)
+- **Ambitious** — 70% grade = success (target = stretch)
+- **Time-boxed** (by end of quarter)
+
+KR types:
+- **Impact metrics:** NRR, activation rate, retention
+- **Leading indicators:** time-to-value, feature adoption
+- **Quality:** p95 latency, error rate, NPS
+
+Mix types — not all impact, not all leading.
+
+✅ Good KR: «Grow weekly active teams from 120 to 180»  
+❌ Bad KR: «Ship checkout redesign» (output, not outcome)
+
+### Step 3 — Ambition Calibration
+
+Stretch goals:
+- **Committed KRs** (must hit) — 100% target. Typically 0.8× ambitious.
+- **Aspirational KRs** — 70% target. Typically 1.3× ambitious.
+
+In B2B SaaS the usual mix: 70% committed + 30% aspirational.
+
+### Step 4 — Initiatives Mapping
+
+For each KR — **which initiatives** (features, projects) drive it? Via `$product-roadmap`.
+
+| KR | Initiative | Expected impact | Timeline |
+|----|------------|:---------------:|----------|
+| KR1 | Onboarding checklist | +15% activation | Week 4-8 |
+
+### Step 5 — Sanity Check
+
+- Does team capacity suffice? (not over-ambitious)
+- Does each Objective have ≥ 2 KRs?
+- Do KRs cover leading + lagging?
+- Conflict check: does achieving one KR not destroy another?
+
+### Step 6 — Review Cadence
+
+- **Weekly:** progress check (burndown style)
+- **Mid-quarter:** re-calibration (adjust committed/aspirational if reality changed)
+- **End-of-quarter:** grade each KR (0-1.0), retro
+
+### Step 7 — Grading
+
+- 0.0-0.3 — failed / didn't prioritize
+- 0.4-0.6 — didn't hit but made progress
+- 0.7 — hit target (healthy for aspirational)
+- 1.0 — fully achieved (if consistently 1.0 → KRs not ambitious enough)
+
+## Validation (Quality Gate)
+
+- [ ] 1-3 Objectives per team / quarter
+- [ ] 2-5 KRs per Objective
+- [ ] KRs measurable (numeric, not subjective)
+- [ ] KRs outcome-focused, not output
+- [ ] Mix leading + lagging
+- [ ] Initiatives mapped to KRs
+- [ ] Capacity-checked
+- [ ] Cascade alignment with company OKR
+- [ ] Review cadence planned
+
+## Handoff
+
+The result is input for:
+- **`product-roadmap`** — initiatives mapped to KRs
+- **`rice-scoring`** — prioritization of initiatives within quarter
+- **Data Analyst** (`$saas-metrics`, `$hypothesis-template`) — how to measure KRs
+
+Format: OKR doc (one page per quarter). Via `$handoff`.
+
+## Anti-patterns
+
+| Error | Why it's bad | How to do it right |
+|-------|-------------|-------------------|
+| Too many Objectives | Focus lost | Max 3 |
+| KR = feature ship | Output, not outcome | Measure customer impact |
+| Vague KR | «Improve X» | Numeric: «X from A to B» |
+| No ambition | Always 1.0 | Aspirational: stretch 70% |
+| Ignoring capacity | Overcommit | Honest capacity check |
+| Copy-paste from company | Not cascaded | «How we contribute to» |
+
+## Template
+
+```markdown
+# OKR Q2 2026 — [Team Name]
+
+## O1: [Inspirational Objective]
+- **KR1.1:** [Metric] from [baseline] to [target] by [date]
+- **KR1.2:** [Metric] from X to Y
+- **KR1.3:** ...
+
+### Initiatives for O1
+| KR | Initiative | Expected impact | Timeline |
+| ... |
+
+## O2: ...
+
+---
+
+## Health Check
+- [ ] Objectives inspire
+- [ ] KRs measurable
+- [ ] Capacity-honest
+- [ ] Cascade checked
+
+## Review Schedule
+- Weekly Tuesday: progress
+- Mid-Q (week 6): re-calibrate
+- End-Q: grade + retro
+```
+
+## Worked Example — TeamFlow Q2 2026 OKRs (Product team, post-Mediator)
+
+**Context:** TeamFlow (B2B SaaS HR-tech, $8M ARR). After Mediator synthesis of visions (Alpha + Beta), the team forms Q2 OKRs for the Product team. Company-level OKRs: grow ARR 50% YoY ($8M → $12M by Q4) + launch AI tier + NRR 105% → 110%.
+
+```markdown
+# OKR Q2 2026 — TeamFlow Product Team
+
+## Company Context (cascade input)
+- **Company O:** Become category-leading People Ops platform by 2027
+- **Company KR:** $12M ARR by Q4 2026 (from $8M), NRR 115%, launch AI tier
+- **Our cascade role:** Drive product-led expansion; enable Enterprise tier; make AI MVP successful
+
+---
+
+## O1: Ship AI 1:1 tier that managers love and buyers pay for
+
+**Ambition:** Aspirational (70% grade = success)
+
+- **KR1.1:** 40 customers (of 200 base) upgraded to AI Team tier ($8/seat) by June 30
+  - Baseline: 0 (tier not existing yet)
+  - Target: 40 accounts = 20% base penetration
+  - Stretch: 60 accounts
+- **KR1.2:** Among upgraded accounts, 65% of active managers use AI summarization weekly (vs zero baseline)
+- **KR1.3:** AI summarization NPS ≥ 50 (vs general TeamFlow NPS 45)
+- **KR1.4:** Action items completion rate lifts from 60% baseline to 75% in AI tier accounts
+
+### Initiatives for O1
+| KR | Initiative | Expected impact | Timeline |
+|----|------------|-----------------|----------|
+| KR1.1 | AI Team tier launch (feature flag gated) | 40 upgrade conversations | Week 4-10 |
+| KR1.1 | GTM campaign (email, webinar, in-app) | +15 upgrades | Week 6-12 |
+| KR1.2 | Onboarding flow for new AI users (checklist + first-success moment) | +25pp adoption | Week 8-10 |
+| KR1.3 | Feedback loop (in-product rating after each summary) | Signal for iteration | Week 6-ongoing |
+| KR1.4 | Automated reminders for AI-extracted action items | +15pp completion | Week 10-12 |
+
+---
+
+## O2: Establish Enterprise tier as credible expansion path
+
+**Ambition:** Committed (100% grade = must hit)
+
+- **KR2.1:** 5 Enterprise tier accounts signed ($50+/seat, min 300 seats each) by June 30
+  - Baseline: 10 Enterprise accounts on legacy pricing
+  - Target: 5 on new Enterprise tier
+- **KR2.2:** Enterprise pilot: 3 design partners live on aggregate dashboard + health score features
+- **KR2.3:** Security review passed: SOC 2 Type II controls documented for AI features (blocker for 2 deals)
+- **KR2.4:** 0 Enterprise contract rejections due to compliance/AI concerns (baseline: 2 in Q1)
+
+### Initiatives for O2
+| KR | Initiative | Expected impact | Timeline |
+|----|------------|-----------------|----------|
+| KR2.1 | Enterprise tier pricing + packaging docs | Sales enablement | Week 2-4 |
+| KR2.2 | Aggregate dashboard MVP (from Beta vision) | 3 design partners engaged | Week 6-12 |
+| KR2.3 | SOC 2 Type II audit for AI tier | Unblock enterprise deals | Week 1-8 |
+| KR2.4 | Compliance FAQ + objection handling deck | Sales reduces objection time | Week 4 |
+
+---
+
+## O3: Build data foundation for 2027 category leadership
+
+**Ambition:** Aspirational (70%)
+
+- **KR3.1:** 1:1 behavioral data pipeline instruments 100% of AI tier events (reliable for downstream analytics)
+- **KR3.2:** Privacy architecture passes external audit (trust layer for future AI features)
+- **KR3.3:** Benchmark engine MVP — 3 industry/segment comparisons live for Enterprise accounts
+
+### Initiatives for O3
+| KR | Initiative | Expected impact | Timeline |
+|----|------------|-----------------|----------|
+| KR3.1 | Event schema + instrumentation (PM + Data) | Data basis for Q3 features | Week 4-8 |
+| KR3.2 | Privacy architecture review + attestation | Moat reinforcement | Week 2-10 |
+| KR3.3 | Benchmark calculation engine | Enterprise delight | Week 10-13 |
+
+---
+
+## Health Check
+- [x] Objectives inspire — all tie to category positioning
+- [x] KRs measurable — all numeric (% / count / NPS / binary)
+- [x] Capacity-honest — 8 eng + 3 PM + 2 design = 13 person-weeks capacity per sprint × 13 sprints = 169 weeks total available; planned initiatives ≈ 120 person-weeks (71% capacity, 29% buffer)
+- [x] Cascade checked — O1 drives Company KR «launch AI tier»; O2 drives «NRR 115%»; O3 drives «category leadership»
+
+## Review Schedule
+- **Weekly Tuesday 10am:** progress burndown + adjustments (15 min per Objective)
+- **Mid-Q (Week 6, May 12):** re-calibrate commitments, move committed↔aspirational if reality shifted
+- **End-Q (June 30):** grade all KRs (0-1.0), retrospective, input for Q3
+
+## Grading Rubric (applied at end-Q)
+
+Example grade for KR1.1 (40 account upgrades):
+- Hit 40 → 1.0 (hit target)
+- Hit 30 → 0.75 (healthy progress, 75% of target)
+- Hit 20 → 0.5 (partial)
+- Hit 10 → 0.25 (minimal)
+- Hit 0 → 0.0
+
+For aspirational KRs, 0.7 = green. For committed KRs, anything below 1.0 triggers retro.
+```
+
+> **OKR lesson:** O1, O2, O3 distribute across Alpha and Beta priorities — O1 is Alpha-weighted (workflow + managers), O2 is Beta-weighted (analytics + buyers), O3 is foundation for both. This prevents OKR-level re-litigation of the Mediator synthesis — the roadmap is now guided by the triple Objective, not by camp politics.