agileflow 4.0.0-alpha.2 → 4.0.0-alpha.21

package/content/plugins/ideation/skills/agileflow-ideation/SKILL.md

@@ -0,0 +1,109 @@
+---
+name: agileflow-ideation
+version: 1.0.0
+category: agileflow/ideation
+description: |
+  Use when the user wants to generate feature ideas, discover
+  opportunities, explore what to build next, or create a product brief.
+  Runs multi-expert brainstorm analyzers across market, UX, growth,
+  integration, and feature gap dimensions.
+triggers:
+  keywords:
+    - what should we build
+    - feature ideas
+    - brainstorm
+    - ideate
+    - what's missing
+    - product ideas
+    - discovery
+    - new features
+    - growth ideas
+    - competitive gaps
+    - what to build next
+  priority: 50
+provides:
+  agents: []
+learns:
+  enabled: true
+  file: _learnings/ideation.yaml
+  maxEntries: 30
+depends:
+  skills: []
+  plugins: [ideation]
+---
+
+# AgileFlow Ideation
+
+Multi-expert feature discovery. Scans the codebase, infers app
+category, and generates prioritized ideas across market positioning,
+UX gaps, growth mechanics, integration opportunities, and missing
+core features.
+
+## When this skill activates
+
+- User asks "what should we build next" or "what's missing"
+- User wants a product brief or feature roadmap input
+- User is doing discovery work before writing stories
+- User wants to find competitive gaps or growth opportunities
+
+## Capabilities
+
+| Command                      | What it analyzes                                        |
+| ---------------------------- | ------------------------------------------------------- |
+| `/agileflow:ideate`          | Full brainstorm — all analyzers                         |
+| `/agileflow:ideate:features` | Missing CRUD, incomplete workflows, half-built features |
+| `/agileflow:ideate:growth`   | Retention hooks, sharing, notifications, activation     |
+| `/agileflow:ideate:discover` | Market-driven gaps, competitive features                |
+| `/agileflow:ideate:new`      | Blank-slate idea generation                             |
+| `/agileflow:ideate:brief`    | Turn ideas into a product brief                         |
+| `/agileflow:ideate:history`  | Review past ideation sessions                           |
+
+## How to guide the user
+
+1. Ask what scope: whole product, a specific feature area, or growth only
+2. Run the appropriate ideation command(s)
+3. After results: filter to 3–5 high-confidence ideas
+4. Offer to turn top ideas into epics via `/agileflow:epic`
+
+## Brainstorm analyzers
+
+The full `/agileflow:ideate` runs 5 analyzers in parallel:
+
+- **Features** — missing core functionality
+- **Growth** — engagement and retention mechanics
+- **Market** — competitive and category-standard features
+- **Integration** — missing third-party connections
+- **UX** — friction points and missing feedback states
+
+Then a consensus agent deduplicates and ranks by confidence × impact.
+
+## Integration
+
+- **agileflow-research** — use before or during ideation to gather competitive intelligence, industry benchmarks, or customer research that sharpens the brainstorm
+- **agileflow-story-writer** — after ideation produces ranked ideas, route the top candidates here to become properly structured user stories with AC
+- **agileflow-epic-planner** — when a cluster of related ideas emerges from ideation, route them to epic-planner to become a sequenced epic with milestones
+- **agileflow-council** — after ideation generates options, convene the council to evaluate strategic trade-offs before committing to a direction
+- **agileflow-planning** — use for scoring and prioritising the ideas from ideation against delivery metrics (velocity, sprint capacity, RICE score)
+- **agileflow-retention** — ideation frequently surfaces engagement and retention opportunities; use retention skill to design the behavioural mechanics for the top ideas
+- **agileflow-ads** — when ideation surfaces a new product angle or audience, use ads to explore how it would position in paid channels
+- **agileflow-seo** — check whether top ideas align with organic search demand before committing to build them
+
+## References
+
+Load these files when you need deeper context for the relevant task:
+
+| File                                          | When to load                                                                              |
+| --------------------------------------------- | ----------------------------------------------------------------------------------------- |
+| `references/feature-prioritization-guide.md`  | Ranking ideas from a brainstorm — RICE, ICE, value/effort matrix, Kano model              |
+| `references/brainstorm-techniques.md`         | Running a brainstorm — HMW, SCAMPER, Crazy 8s, pre-mortem, dot voting                     |
+| `references/user-story-patterns.md`           | Turning ideas into stories — JTBD format, persona templates, acceptance criteria patterns |
+| `references/competitive-analysis-template.md` | Analyzing competitors — comparison dimensions, scoring, gap identification                |
+
+## Workflows
+
+Follow these step-by-step when the user initiates the matching action:
+
+| File                    | When to follow                                                                                     |
+| ----------------------- | -------------------------------------------------------------------------------------------------- |
+| `workflows/ideate.md`   | User wants to generate feature ideas — runs multi-expert brainstorm and surfaces top opportunities |
+| `workflows/features.md` | User wants a focused feature brainstorm for a specific area of the product                         |

package/content/plugins/ideation/skills/agileflow-ideation/references/brainstorm-techniques.md

@@ -0,0 +1,138 @@
+# Brainstorm Techniques
+
+**Load this when:** facilitating a brainstorm, generating feature ideas,
+or helping a team get unstuck on a problem.
+
+## Technique selection guide
+
+| Goal                       | Best technique                   |
+| -------------------------- | -------------------------------- |
+| Generate many ideas fast   | Brainwriting / silent brainstorm |
+| Find unexpected angles     | SCAMPER, "How might we"          |
+| Understand user problems   | Jobs-to-be-done interviews       |
+| Explore adjacent markets   | Analogical thinking              |
+| Challenge assumptions      | Pre-mortem, "Kill the company"   |
+| Prioritize from large list | Dot voting, RICE                 |
+
+## How Might We (HMW)
+
+Reframes problems as opportunities. Takes a user pain and makes it a design prompt.
+
+**Formula:** "How might we [remove obstacle / improve experience / enable behavior]?"
+
+**Examples:**
+
+- Pain: "Users don't know if their order shipped" → HMW: "How might we give users confidence their order is on its way without them having to check?"
+- Pain: "Onboarding takes 20 minutes" → HMW: "How might we get users to their first 'aha' in under 3 minutes?"
+- Pain: "Reports are confusing" → HMW: "How might we make our reports understandable to a non-technical manager?"
+
+HMW questions shouldn't imply the solution (avoid "How might we send more emails?").
+
+## SCAMPER
+
+Structured creativity — force new angles by applying transformations:
+
+| Letter                  | Prompt                           | Applied to a feature                                 |
+| ----------------------- | -------------------------------- | ---------------------------------------------------- |
+| **S**ubstitute          | Replace a component              | Replace email notifications with push                |
+| **C**ombine             | Merge with something else        | Combine onboarding with first use                    |
+| **A**dapt               | Borrow from another domain       | Adapt Spotify's Discover Weekly to surface old tasks |
+| **M**odify / Magnify    | Make bigger, smaller, faster     | Make the dashboard load instantly (edge caching)     |
+| **P**ut to other uses   | New context for existing feature | Use billing emails as a re-engagement channel        |
+| **E**liminate           | Remove a step                    | Remove the confirmation email entirely               |
+| **R**everse / Rearrange | Flip the order                   | Ask for payment before account creation              |
+
+## Crazy 8s
+
+8 ideas in 8 minutes. No judgment, no refining — volume over quality.
+
+**Format:**
+
+1. Set a timer for 8 minutes
+2. Fold a sheet into 8 panels (or use 8 sticky notes)
+3. Sketch one idea per panel — 1 minute each
+4. Review after — usually 1-2 are surprisingly strong
+
+Works best for UI/UX problems. Forces quantity, breaks perfectionism.
+
+## Pre-mortem
+
+Assume the project failed. Work backward to find why.
+
+**Script:**
+
+1. "Imagine it's 12 months from now and this feature completely failed."
+2. "What went wrong? List every possible cause."
+3. Review the list — which risks can be mitigated now?
+
+More effective than forward-looking risk analysis because it activates different thinking.
+
+## Analogical thinking
+
+Find solutions in adjacent industries:
+
+| Your problem                     | Analogous domain                          | Insight                                          |
+| -------------------------------- | ----------------------------------------- | ------------------------------------------------ |
+| Reduce checkout abandonment      | Hotel industry: hold room without payment | Allow "save my cart" with intent signal          |
+| Onboarding takes too long        | Airlines: pre-check                       | Pre-fill everything available before user starts |
+| Engagement drops after day 7     | Fitness apps: streak mechanics            | Day-7 milestone + visible progress streak        |
+| Support tickets spike on Mondays | Restaurants: Sunday prep                  | "Monday preview" email sent Fridays              |
+
+**Process:**
+
+1. Abstract your problem to its essence: "How do we maintain engagement over time?"
+2. Ask: "Which industries are exceptional at this?"
+3. Study their approach
+4. Translate back to your context
+
+## "Kill the company" exercise
+
+Forces teams to find their biggest vulnerabilities:
+
+1. "A competitor just put us out of business. What did they do?"
+2. "A regulation just killed our product. What was it?"
+3. "Our biggest customer just left. What drove them away?"
+
+The answers reveal your defensive priorities and often surface product gaps worth building against.
+
+## Brainwriting (silent brainstorm)
+
+Better than spoken brainstorms for introverts and reducing groupthink:
+
+1. Each person writes 3 ideas (5 min, silent)
+2. Pass papers clockwise
+3. Read neighbor's ideas, write 3 more (inspired or independent) — 5 min
+4. Repeat 3-4 rounds
+5. Collect all, group by theme, dot vote
+
+**Why it's better than talking:** No anchoring on first idea, equal participation, no social pressure.
+
+## Dot voting
+
+Quick prioritization from a large list:
+
+1. Put all ideas on a board (sticky notes or digital)
+2. Each person gets 3-5 dots (votes)
+3. Vote silently — can cluster or spread
+4. Count dots, discuss top items
+
+**Bias warning:** People vote for ideas they already know. Require voters to read all ideas before voting.
+
+## Feature analogy generation
+
+Prompt for generating adjacent feature ideas:
+
+```
+Our product does [X].
+Users also need [related task].
+What feature would bridge [X] to [related task]?
+```
+
+Example:
+
+```
+Our product does project management.
+Users also need to communicate with clients.
+What feature would bridge project management to client communication?
+→ Client-visible comments, shareable milestones, client portal
+```

package/content/plugins/ideation/skills/agileflow-ideation/references/competitive-analysis-template.md

@@ -0,0 +1,148 @@
+# Competitive Analysis Template
+
+**Load this when:** Researching competitors, positioning a product, or evaluating market gaps.
+
+## Competitor Profile Card
+
+```markdown
+## [Competitor Name]
+
+**Category:** Direct competitor | Indirect | Emerging threat | Adjacent
+**URL:**
+**Pricing:** [Free tier? Pricing model — per seat / usage / flat]
+**Target customer:** [ICP description]
+**Core value prop (their words):** [Quote from their homepage]
+**Our assessment of their actual value prop:** [What do customers actually love?]
+
+### Strengths
+
+-
+-
+
+### Weaknesses
+
+-
+-
+
+### Notable features we don't have
+
+-
+
+### Features they lack that we have
+
+-
+
+### Pricing vs. ours
+
+[Table or prose comparison]
+```
+
+---
+
+## Comparison Dimensions
+
+### Core product dimensions
+
+| Dimension             | Us  | Competitor A | Competitor B | Notes |
+| --------------------- | --- | ------------ | ------------ | ----- |
+| Primary use case      |     |              |              |       |
+| Target segment        |     |              |              |       |
+| Onboarding time       |     |              |              |       |
+| Integration ecosystem |     |              |              |       |
+| Mobile support        |     |              |              |       |
+| Offline support       |     |              |              |       |
+| API availability      |     |              |              |       |
+| SSO / enterprise auth |     |              |              |       |
+| Data export           |     |              |              |       |
+| Audit logs            |     |              |              |       |
+
+### Pricing model comparison
+
+| Dimension           | Us  | Competitor A | Competitor B |
+| ------------------- | --- | ------------ | ------------ |
+| Free tier           |     |              |              |
+| Entry price         |     |              |              |
+| Scale trigger       |     |              |              |
+| Enterprise / custom |     |              |              |
+| Per-seat or usage   |     |              |              |
+| Contract required   |     |              |              |
+
+---
+
+## Positioning Scoring Matrix
+
+Rate each on 1–5 scale for target customer:
+
+| Dimension                 | Us  | Comp A | Comp B | Weight |
+| ------------------------- | --- | ------ | ------ | ------ |
+| Ease of onboarding        |     |        |        | 20%    |
+| Feature depth             |     |        |        | 15%    |
+| Performance / reliability |     |        |        | 20%    |
+| Pricing competitiveness   |     |        |        | 15%    |
+| Integration breadth       |     |        |        | 10%    |
+| Support quality           |     |        |        | 10%    |
+| Documentation quality     |     |        |        | 10%    |
+
+**Weighted score = sum(score × weight)**
+
+---
+
+## Market Gap Analysis
+
+After filling the comparison matrix, identify:
+
+```
+Gaps (they have it, we don't):
+1.
+2.
+
+Opportunities (nobody has it, customer needs it):
+1.
+2.
+
+Advantages (we have it, they don't):
+1.
+2.
+
+Parity features (table stakes, must-have):
+1.
+2.
+```
+
+---
+
+## Win/Loss Signals to Track
+
+| Signal                         | Where to find it                             |
+| ------------------------------ | -------------------------------------------- |
+| Why customers chose us         | Sales call notes, onboarding surveys         |
+| Why customers chose competitor | Lost deal notes, churned user interviews     |
+| Public reviews                 | G2, Capterra, Trustpilot, App Store          |
+| Feature requests               | Support tickets, forum posts, Twitter/Reddit |
+| Pricing sensitivity            | Sales objections, deal slippage patterns     |
+| Integration gaps               | Support requests for missing integrations    |
+
+---
+
+## Competitive Intelligence Sources
+
+| Source                    | Type                    | Update frequency |
+| ------------------------- | ----------------------- | ---------------- |
+| Their changelog / blog    | Feature releases        | Weekly check     |
+| G2, Capterra reviews      | User sentiment          | Monthly          |
+| LinkedIn job postings     | Strategic direction     | Quarterly        |
+| Their pricing page        | Pricing changes         | Monthly          |
+| Twitter/Reddit mentions   | Sentiment, feature gaps | Weekly           |
+| SEO keyword gaps          | Content/product focus   | Quarterly        |
+| Crunchbase / funding news | Company trajectory      | On news          |
+
+---
+
+## Competitive Response Framework
+
+| Scenario                          | Response                                                                                |
+| --------------------------------- | --------------------------------------------------------------------------------------- |
+| Competitor adds a feature we lack | Assess: parity vs. differentiation. Add if table stakes, ignore if not our ICP.         |
+| Competitor cuts price             | Don't race to bottom. Compete on value. Quantify ROI difference.                        |
+| New competitor enters market      | Document, monitor, don't panic. Validate with customers if it changes buying behavior.  |
+| Competitor gets acquired          | Research acquirer's intent. Could mean faster growth (threat) or neglect (opportunity). |

package/content/plugins/ideation/skills/agileflow-ideation/references/feature-prioritization-guide.md

@@ -0,0 +1,147 @@
+# Feature Prioritization Guide
+
+**Load this when:** prioritizing features from a backlog, choosing between
+competing ideas, or building a scoring model for a brainstorm.
+
+## RICE scoring
+
+**R**each × **I**mpact × **C**onfidence ÷ **E**ffort
+
+| Factor         | What it measures           | Scale                                              |
+| -------------- | -------------------------- | -------------------------------------------------- |
+| **Reach**      | Users affected per quarter | Count (e.g., 500 users)                            |
+| **Impact**     | Effect on the metric       | 3=massive, 2=high, 1=medium, 0.5=low, 0.25=minimal |
+| **Confidence** | How sure you are           | 100%=high, 80%=medium, 50%=low                     |
+| **Effort**     | Person-months              | Estimate directly                                  |
+
+```
+RICE = (Reach × Impact × Confidence%) / Effort
+```
+
+Higher = higher priority. Compare features to each other, not absolute values.
+
+**Example:**
+
+```
+Feature: Email digest notifications
+Reach: 2,000 users/quarter
+Impact: 1 (medium — improves retention, not conversion)
+Confidence: 80%
+Effort: 0.5 person-months
+
+RICE = (2000 × 1 × 0.8) / 0.5 = 3,200
+```
+
+## ICE scoring (faster, less precise)
+
+**I**mpact × **C**onfidence × **E**ase
+
+All on a 1-10 scale. Ease = 10 minus effort (10=trivial, 1=massive).
+
+Good for quick team alignment or solo backlog grooming.
+
+## Value vs effort matrix
+
+```
+HIGH IMPACT
+    │  Quick wins ⭐  │  Major bets 🔮  │
+    │  (do first)     │  (plan carefully)│
+    │─────────────────┼─────────────────│
+    │  Maybe later    │  Avoid ❌       │
+    │  (deprioritize) │  (cut)          │
+LOW IMPACT
+         LOW EFFORT        HIGH EFFORT
+```
+
+**Quick wins**: Ship these first. High return for low investment.
+**Major bets**: Worth doing but need careful planning. Sequence after quick wins.
+**Maybe later**: Low impact, low effort. Do only if bandwidth allows.
+**Avoid**: Cut from backlog. High effort, low return.
+
+## Kano model — feature delight mapping
+
+| Category        | User reaction if absent | User reaction if present | Example                   |
+| --------------- | ----------------------- | ------------------------ | ------------------------- |
+| **Basic**       | Angry/dissatisfied      | Indifferent              | App doesn't crash         |
+| **Performance** | Less satisfied          | More satisfied           | Faster load time          |
+| **Excitement**  | Not missed              | Delighted                | Dark mode, magic features |
+
+**Strategy:**
+
+- Basic features: fix first, never ship without
+- Performance features: invest proportionally to impact
+- Excitement features: differentiators, keep some for launch moments
+
+## Jobs-to-be-done framing
+
+Before scoring, define the job:
+
+```
+When [situation], I want to [motivation], so I can [expected outcome].
+```
+
+**Example:**
+
+```
+When I receive a new order at 2am, I want to get a notification on my phone,
+so I can decide whether to wake my team for urgent fulfillment.
+```
+
+Features that serve a well-defined JTBD rank higher — they have a clear user need.
+Features without a JTBD are often solution-first thinking (someone wanted to build it,
+not someone needed it).
+
+## Dependency sequencing
+
+Some features must come before others:
+
+```
+OAuth integration
+  └─ Social login (requires OAuth)
+       └─ "Continue with Google" button (requires social login)
+
+DB schema: user preferences table
+  └─ API: save/load preferences
+       └─ UI: preferences panel
+```
+
+Always map blockers before scoring — a high-RICE feature blocked by 3 other features
+effectively has the sum of all their effort.
+
+## Balancing the portfolio
+
+Don't optimize only for highest RICE. Balance across:
+
+| Category       | Target mix | Purpose                     |
+| -------------- | ---------- | --------------------------- |
+| Quick wins     | 30-40%     | Momentum, visible progress  |
+| Major bets     | 20-30%     | Competitive differentiation |
+| Debt/stability | 15-20%     | Sustain velocity            |
+| Experiments    | 10-15%     | Learning, discovery         |
+
+A backlog with only quick wins stagnates. A backlog with only major bets ships nothing.
+
+## When to cut a feature
+
+Cut (or park for 6+ months) when:
+
+- RICE score is in the bottom 25% of the backlog
+- No clear JTBD — "it would be nice to have"
+- Blocked by 3+ other features with unknown timelines
+- Similar feature exists and usage is <5%
+- Team can't explain who specifically would use it
+
+**Parking ≠ deleting.** Move to "icebox" with the original request and date.
+Revisit quarterly. If still unasked for, delete.
+
+## Communicating prioritization decisions
+
+When a stakeholder's feature gets deprioritized:
+
+1. Acknowledge the request by name
+2. Show where it landed in scoring (transparent)
+3. Explain what beat it and why
+4. Give a realistic timeline (or honest "not this quarter")
+5. Offer an alternative if one exists
+
+Never just say "we'll get to it." Give a date or a condition: "we'll revisit when X is shipped."

package/content/plugins/ideation/skills/agileflow-ideation/references/user-story-patterns.md

@@ -0,0 +1,152 @@
+# User Story Patterns
+
+**Load this when:** Writing user stories, defining acceptance criteria, or structuring feature requirements.
+
+## Standard Story Formats
+
+### Classic format
+
+```
+As a [persona/role],
+I want to [action/goal],
+So that [benefit/outcome].
+```
+
+### JTBD (Jobs-to-Be-Done) format
+
+Better for product strategy and feature discovery:
+
+```
+When [situation/trigger],
+I want to [motivation/goal],
+So I can [expected outcome].
+```
+
+Example:
+
+> When I'm reviewing a PR late at night,
+> I want to see a summary of what changed and why,
+> So I can give a meaningful review without reading every line.
+
+### Outcome-first format
+
+Better for roadmap items and stakeholder communication:
+
+```
+[Persona] needs to [accomplish outcome]
+because [root cause / current friction].
+```
+
+---
+
+## Persona Templates
+
+### Lightweight persona card
+
+```
+Name: [Role name — "Alex the Developer", not a real person name]
+Context: [Where they are in the workflow]
+Goal: [What they're trying to accomplish right now]
+Frustration: [What currently gets in the way]
+Success looks like: [What they'd say if it worked perfectly]
+```
+
+### Persona tiers for B2B products
+
+| Tier     | Role                             | Decision power   | Story focus                |
+| -------- | -------------------------------- | ---------------- | -------------------------- |
+| Champion | Power user / day-to-day operator | Feature adoption | Workflow efficiency, depth |
+| Buyer    | Manager / Director               | Budget approval  | ROI, reporting, compliance |
+| End user | Occasional user                  | Usage            | Simplicity, onboarding     |
+| Admin    | IT / DevOps                      | Setup + security | Config, SSO, audit logs    |
+
+---
+
+## Acceptance Criteria Patterns
+
+### Given/When/Then (BDD format)
+
+Best for testable, specific behavior:
+
+```
+Given [precondition / context]
+When [action / event]
+Then [expected outcome]
+  And [additional outcome]
+```
+
+### Checklist format
+
+Best for UI/UX stories with multiple states:
+
+```
+Acceptance criteria:
+- [ ] [Condition 1 — observable, verifiable]
+- [ ] [Condition 2]
+- [ ] [Error state handled: ...]
+- [ ] [Empty state handled: ...]
+- [ ] [Loading state handled: ...]
+```
+
+### Constraint format
+
+Best for non-functional requirements:
+
+```
+The system must [behavior] [qualifier].
+Example: "The search results must appear within 300ms for queries under 10k records."
+```
+
+---
+
+## Story Sizing Signals
+
+| Story complexity signal         | Suggested action               |
+| ------------------------------- | ------------------------------ |
+| More than 3 acceptance criteria | Consider splitting             |
+| Involves >2 system boundaries   | Split by boundary              |
+| Requires new data model design  | Spike story first              |
+| "And also..." in story title    | Split on the "and also"        |
+| Multiple distinct user types    | One story per persona          |
+| Unclear success metric          | Write definition of done first |
+
+---
+
+## Story Splitting Patterns
+
+| Pattern           | How to split                    | Example                                          |
+| ----------------- | ------------------------------- | ------------------------------------------------ | ------------------------------------------- |
+| By workflow step  | One story per step              | "View cart" / "Edit cart" / "Checkout"           |
+| By data variation | One story per type              | "Import CSV" / "Import JSON"                     |
+| >                 | By happy/unhappy path           | One for success, one for error                   | "Send notification" / "Handle send failure" |
+| By user type      | One story per persona           | "Admin dashboard" / "Read-only viewer dashboard" |
+| By complexity     | Simple first, edge cases later  | "Basic search" → "Search with filters"           |
+| CRUD split        | Create / Read / Update / Delete | Four separate stories                            |
+
+---
+
+## Definition of Done (standard checklist)
+
+Adapt per team but include:
+
+- [ ] Acceptance criteria all met
+- [ ] Unit tests written and passing
+- [ ] Code reviewed and approved
+- [ ] No new linting errors
+- [ ] Deployed to staging, smoke-tested
+- [ ] Documentation updated (if applicable)
+- [ ] Feature flag configured (if applicable)
+- [ ] Accessibility requirements met (if UI)
+
+---
+
+## Anti-Patterns to Avoid
+
+| Anti-pattern                            | Problem                            | Fix                     |
+| --------------------------------------- | ---------------------------------- | ----------------------- |
+| "As a user, I want to click the button" | Describes implementation, not need | Reframe around outcome  |
+| "As a developer, I want to refactor X"  | Tech tasks aren't stories          | Use a task/chore ticket |
+| No acceptance criteria                  | Untestable; done when?             | Add at least 2 criteria |
+| Acceptance criteria as a design spec    | Too prescriptive                   | Test outcomes, not UI   |
+| Epic-sized story                        | Takes >1 sprint                    | Split it                |
+| Vague benefit ("so that it's better")   | No value validation possible       | Make benefit measurable |