@qball-inc/the-bulwark 1.0.0

package/skills/product-ideation/SKILL.md

@@ -0,0 +1,372 @@
+---
+name: product-ideation
+description: Evaluates product ideas through a six-agent research pipeline producing BUY/HOLD/SELL recommendations backed by market analysis, competitive intelligence, and segment targeting. Use when assessing product viability, researching markets for new ideas, or deciding whether to pursue a product concept.
+user-invocable: true
+argument-hint: "<idea-description> | --doc <path-to-idea-file>"
+skills:
+  - subagent-prompting
+---
+
+# Product Ideation
+
+Evaluates a product idea through a structured six-agent research pipeline and delivers a BUY/HOLD/SELL recommendation. The orchestrator interviews the user to clarify the idea, then spawns dedicated sub-agents for validation, market research, competitive analysis, pattern documentation, segment analysis, and final strategy. All analysis is grounded in web research and structured frameworks.
+
+---
+
+## When to Use This Skill
+
+**Load this skill when the user request matches ANY of these patterns:**
+
+| Trigger Pattern | Example User Request |
+|-----------------|----------------------|
+| Idea evaluation | "Is this product idea worth pursuing?", "Evaluate my idea: X" |
+| Market analysis | "Research the market for X", "Should I build X?" |
+| Investment-style assessment | "Give me a BUY/HOLD/SELL on this idea", "Validate my product concept" |
+| Pre-build research | "Before I start building X, help me understand the landscape" |
+
+**DO NOT use for:**
+- Evaluating existing live products or companies (use a financial analysis skill instead)
+- Technical architecture decisions for an already-committed product (the idea stage has passed)
+- Competitive analysis in isolation without a product idea as the anchor
+
+---
+
+## Dependencies
+
+| Category | Files | Requirement | When to Load |
+|----------|-------|-------------|--------------|
+| **Analysis frameworks** | `references/analysis-frameworks.md` | **REQUIRED** | Include in sub-agent prompts at Stage 1+ |
+| **Final report template** | `templates/final-report.md` | **REQUIRED** | Include in strategist sub-agent prompt |
+| **Diagnostics schema** | `templates/diagnostic-output.yaml` | **REQUIRED** | Write at pipeline completion |
+| **Prompting** | `subagent-prompting` skill | **REQUIRED** | Load before spawning any sub-agent |
+| **Idea validator** | `.claude/agents/product-ideation-idea-validator.md` | **REQUIRED** | Spawned at Stage 1 (parallel) |
+| **Market researcher** | `.claude/agents/product-ideation-market-researcher.md` | **REQUIRED** | Spawned at Stage 1 (parallel) |
+| **Competitive analyzer** | `.claude/agents/product-ideation-competitive-analyzer.md` | **REQUIRED** | Spawned at Stage 2 (sequential) |
+| **Pattern documenter** | `.claude/agents/product-ideation-pattern-documenter.md` | **REQUIRED** | Spawned at Stage 3 (parallel) |
+| **Segment analyzer** | `.claude/agents/product-ideation-segment-analyzer.md` | **REQUIRED** | Spawned at Stage 3 (parallel) |
+| **Strategist** | `.claude/agents/product-ideation-strategist.md` | **REQUIRED** | Spawned at Stage 4 (sequential) |
+
+---
+
+## Usage
+
+```
+/product-ideation <idea-description>
+/product-ideation --doc <path-to-idea-file>
+```
+
+**Arguments:**
+- `<idea-description>` — Free-text description of the product idea
+- `--doc <path>` — Path to a file containing the idea description
+
+**Examples:**
+- `/product-ideation a mobile app that lets freelancers track client payments with automated invoice reminders`
+- `/product-ideation --doc plans/ideas/saas-invoicing-tool.md`
+
+---
+
+## Pre-Flight Gate (BLOCKING)
+
+**STOP. Before ANY work, you MUST acknowledge what this skill requires.**
+
+This skill uses a **six-agent pipeline with parallel and sequential stages**. You are the orchestrator. You coordinate, read outputs, and spawn agents — you do NOT perform their analysis yourself.
+
+### What You MUST Do
+
+1. Parse input and conduct a focused interview (Stage 0) — at most 2 AskUserQuestion rounds
+2. Read `references/analysis-frameworks.md` and `templates/final-report.md` before spawning agents
+3. Spawn Stage 1 agents (idea-validator AND market-researcher) in the same message as parallel Tasks
+4. Wait for both Stage 1 agents to complete before spawning Stage 2
+5. Spawn Stage 3 agents (pattern-documenter AND segment-analyzer) in the same message as parallel Tasks
+6. Wait for all prior outputs before spawning the strategist
+7. Read the final strategy report from `$PROJECT_DIR/logs/` and present its contents to the user
+8. Write diagnostic YAML at completion
+
+### What You MUST NOT Do
+
+- Do NOT perform market research, validation, or competitive analysis yourself — spawn the agents
+- Do NOT skip the interview — the idea must be clear before spawning agents
+- Do NOT spawn agents with `run_in_background: true`
+- Do NOT proceed past Stage 1 without confirming both parallel agents completed
+- Do NOT return to user before the strategist has written its output to `$PROJECT_DIR/logs/`
+
+**Anti-thought traps:**
+- "I already know this market, I can skip the research agents" — STOP. Spawn the agents.
+- "The idea is simple, I don't need to interview" — STOP. Conduct at least one AskUserQuestion round.
+- "I can summarize the recommendation without reading the strategist output" — STOP. Read the log file.
+
+---
+
+## Pipeline
+
+```fsharp
+// product-ideation pipeline
+Stage0_Interview(args)
+|> [Stage1_IdeaValidator(idea),         // product-ideation-idea-validator — feasibility + timing
+    Stage1_MarketResearcher(idea)]       // product-ideation-market-researcher — market size + trends
+|> Stage2_CompetitiveAnalyzer(market_research_output)  // product-ideation-competitive-analyzer
+|> [Stage3_PatternDocumenter(competitive_output),       // product-ideation-pattern-documenter
+    Stage3_SegmentAnalyzer(competitive_output)]         // product-ideation-segment-analyzer
+|> Stage4_Strategist(all_outputs)       // product-ideation-strategist — BUY/HOLD/SELL
+|> Stage5_Diagnostics(pipeline_results)
+```
+
+---
+
+## Stage Definitions
+
+### Stage 0: Interview (Orchestrator)
+
+```
+Stage 0: Interview
+├── Parse arguments:
+│   ├── If --doc <path>: Read the file, extract idea description
+│   └── If free-text: Use as idea description directly
+├── Load references/analysis-frameworks.md
+├── Load templates/final-report.md
+├── AskUserQuestion (Round 1 — REQUIRED):
+│   ├── Q1: What problem does this product solve, and for whom?
+│   ├── Q2: What is your rough idea of the business model (subscription, one-time, marketplace, etc.)?
+│   ├── Q3: Are there any specific markets or geographies to focus on?
+│   └── Q4: Do you have any known competitors in mind, or is this a blank slate?
+├── If answers are ambiguous or incomplete:
+│   └── AskUserQuestion (Round 2 — CONDITIONAL, max 1 additional round):
+│       └── Ask 1-2 targeted follow-ups based on gaps in Round 1 answers
+├── Synthesize: Combine original idea + interview answers into a complete idea brief
+└── Token budget check (warn if >20% consumed before agents start)
+```
+
+**Interview behavior**: Maximum 2 AskUserQuestion rounds. Present all Round 1 questions together. Do NOT ask questions one at a time.
+
+### Stage 1: Parallel Batch 1 — Idea Validator + Market Researcher
+
+Spawn both agents in the same message. Do NOT wait for one before spawning the other.
+
+**Idea Validator prompt structure:**
+
+```
+GOAL: Assess whether the product idea has sufficient merit to warrant deep market research.
+Evaluate technical feasibility, market timing, problem-solution fit, and uniqueness.
+
+CONSTRAINTS:
+- Use WebSearch to check for prior art and existing solutions
+- Produce a clear PASS/CONDITIONAL/FAIL verdict with reasoning
+- Write output to $PROJECT_DIR/logs/product-ideation-idea-validation-{YYYYMMDD-HHMMSS}.md
+- Do not exceed 500 words in the main report
+
+CONTEXT:
+Idea brief: {synthesized idea brief from Stage 0}
+Frameworks reference: {path to references/analysis-frameworks.md}
+
+OUTPUT:
+Write report to $PROJECT_DIR/logs/product-ideation-idea-validation-{YYYYMMDD-HHMMSS}.md
+Return summary to orchestrator: verdict, top 3 strengths, top 3 concerns, report path
+```
+
+Spawn: `Task(subagent_type="product-ideation-idea-validator", prompt=...)`
+
+**Market Researcher prompt structure:**
+
+```
+GOAL: Research the market landscape for this product idea. Produce an evidence-based
+assessment of market size, growth trajectory, key players, regulatory context, and
+technology trends. Ground all claims in web research.
+
+CONSTRAINTS:
+- Use WebSearch and WebFetch extensively — minimum 5 distinct sources
+- Quantify market size where possible (TAM, growth rate, CAGR)
+- Note recency of data — prefer sources from 2023 or later
+- Write output to $PROJECT_DIR/logs/product-ideation-market-research-{YYYYMMDD-HHMMSS}.md
+
+CONTEXT:
+Idea brief: {synthesized idea brief from Stage 0}
+Frameworks reference: {path to references/analysis-frameworks.md} — use PESTLE and TAM/SAM/SOM
+
+OUTPUT:
+Write report to $PROJECT_DIR/logs/product-ideation-market-research-{YYYYMMDD-HHMMSS}.md
+Return summary to orchestrator: market size estimate, growth trend, top 3 key players, report path
+```
+
+Spawn: `Task(subagent_type="product-ideation-market-researcher", prompt=...)`
+
+After both agents return: verify both log files exist. If either is missing, re-spawn that agent once.
+
+### Stage 2: Sequential — Competitive Analyzer
+
+Read the market research log file path from Stage 1 output before constructing this prompt.
+
+```
+GOAL: Perform competitive intelligence analysis. Identify direct competitors, indirect
+competitors, and failed competitors. Assess their positioning, pricing, strengths,
+weaknesses, and the reasons for failure in failed cases.
+
+CONSTRAINTS:
+- Read market research output from: {market_research_log_path}
+- Use WebSearch to discover competitors not mentioned in market research
+- Include at least 3 direct competitors and 2 failed competitors
+- Write output to $PROJECT_DIR/logs/product-ideation-competitive-analysis-{YYYYMMDD-HHMMSS}.md
+
+CONTEXT:
+Idea brief: {synthesized idea brief from Stage 0}
+Market research: {market_research_log_path}
+Frameworks reference: {path to references/analysis-frameworks.md} — use Porter's Five Forces
+
+OUTPUT:
+Write report to $PROJECT_DIR/logs/product-ideation-competitive-analysis-{YYYYMMDD-HHMMSS}.md
+Return summary to orchestrator: top 3 competitors, market gap identified (yes/no), report path
+```
+
+Spawn: `Task(subagent_type="product-ideation-competitive-analyzer", prompt=...)`
+
+### Stage 3: Parallel Batch 2 — Pattern Documenter + Segment Analyzer
+
+Read the competitive analysis log file path from Stage 2 output before spawning. Spawn both agents in the same message.
+
+**Pattern Documenter prompt structure:**
+
+```
+GOAL: Document success patterns and failure patterns from the competitive landscape.
+Produce competitor profiles, identify opportunity gaps, and extract repeatable lessons.
+
+CONSTRAINTS:
+- Read competitive analysis from: {competitive_analysis_log_path}
+- May use WebSearch to deepen understanding of specific competitor trajectories
+- Document at least 3 success patterns and 3 failure patterns
+- Write output to $PROJECT_DIR/logs/product-ideation-patterns-{YYYYMMDD-HHMMSS}.md
+
+CONTEXT:
+Idea brief: {synthesized idea brief from Stage 0}
+Competitive analysis: {competitive_analysis_log_path}
+Frameworks reference: {path to references/analysis-frameworks.md}
+
+OUTPUT:
+Write report to $PROJECT_DIR/logs/product-ideation-patterns-{YYYYMMDD-HHMMSS}.md
+Return summary to orchestrator: top success pattern, top failure risk, biggest opportunity gap, report path
+```
+
+Spawn: `Task(subagent_type="product-ideation-pattern-documenter", prompt=...)`
+
+**Segment Analyzer prompt structure:**
+
+```
+GOAL: Identify and profile target user segments for this product idea. Determine who
+will buy it, why, at what price, and in what volume. Produce TAM/SAM/SOM estimates
+and at least 2 detailed user personas.
+
+CONSTRAINTS:
+- Read competitive analysis from: {competitive_analysis_log_path}
+- Use WebSearch to validate segment size claims
+- Produce at least 2 distinct user personas with willingness-to-pay estimates
+- Write output to $PROJECT_DIR/logs/product-ideation-segments-{YYYYMMDD-HHMMSS}.md
+
+CONTEXT:
+Idea brief: {synthesized idea brief from Stage 0}
+Competitive analysis: {competitive_analysis_log_path}
+Frameworks reference: {path to references/analysis-frameworks.md} — use Jobs-to-be-Done + TAM/SAM/SOM
+
+OUTPUT:
+Write report to $PROJECT_DIR/logs/product-ideation-segments-{YYYYMMDD-HHMMSS}.md
+Return summary to orchestrator: primary segment, estimated TAM, top persona name, report path
+```
+
+Spawn: `Task(subagent_type="product-ideation-segment-analyzer", prompt=...)`
+
+After both agents return: verify both log files exist.
+
+### Stage 4: Sequential — Strategist
+
+Read all four prior log file paths before constructing this prompt.
+
+```
+GOAL: Synthesize all research outputs into a definitive BUY/HOLD/SELL recommendation.
+Produce the final report using the provided template. Support the recommendation with
+confidence level, strategic rationale, risk factors, and next steps.
+
+CONSTRAINTS:
+- Read ALL of the following before writing:
+    {idea_validation_log_path}
+    {market_research_log_path}
+    {competitive_analysis_log_path}
+    {patterns_log_path}
+    {segments_log_path}
+- Use the final-report template at: {path to templates/final-report.md}
+- Recommendation MUST be exactly one of: BUY, HOLD, or SELL
+- Confidence MUST be exactly one of: High, Medium, or Low
+- Write output to $PROJECT_DIR/logs/product-ideation-strategy-{YYYYMMDD-HHMMSS}.md
+
+CONTEXT:
+Idea brief: {synthesized idea brief from Stage 0}
+All prior outputs: see CONSTRAINTS for paths
+BUY/HOLD/SELL criteria: {path to references/analysis-frameworks.md} — see Recommendation Thresholds section
+Final report template: {path to templates/final-report.md}
+
+OUTPUT:
+Write final report to $PROJECT_DIR/logs/product-ideation-strategy-{YYYYMMDD-HHMMSS}.md
+Return summary to orchestrator: recommendation, confidence level, one-sentence rationale, report path
+```
+
+Spawn: `Task(subagent_type="product-ideation-strategist", prompt=...)`
+
+After agent returns: read the full strategy report from the log path. Present its contents to the user verbatim (do not summarize — this is the deliverable).
+
+### Stage 5: Diagnostics (REQUIRED)
+
+**MANDATORY**: Write diagnostic YAML after every invocation. Cannot be skipped.
+
+```
+Write to: $PROJECT_DIR/logs/diagnostics/product-ideation-{YYYYMMDD-HHMMSS}.yaml
+Use schema from: templates/diagnostic-output.yaml
+
+Include:
+- Input type and value
+- Interview rounds conducted
+- Each pipeline stage: success | failed | skipped
+- Final recommendation and confidence
+- All log file paths
+- Outcome: success | partial | failure
+```
+
+---
+
+## Error Handling
+
+| Scenario | Action |
+|----------|--------|
+| Agent returns empty output | Re-spawn once with reinforced instructions. If still empty, mark stage as failed in diagnostics, skip to next stage. |
+| Idea validation returns FAIL | Present the FAIL verdict to user. AskUserQuestion: "The idea validator flagged significant concerns. Proceed anyway, or refine the idea?" |
+| Market researcher finds no data | Note data gap in final report. Strategist must lower confidence to Low if market size cannot be estimated. |
+| Both Stage 1 agents fail | STOP. Inform user: "Research pipeline failed at Stage 1. Please retry with a more specific idea description." |
+| Log file missing after agent completes | Glob for similar filenames in `$PROJECT_DIR/logs/`. If found, use it. If not, re-spawn agent. |
+| Token budget >60% at Stage 3 | Warn user. Spawn agents sequentially instead of in parallel to conserve context. |
+| Token budget >80% at any point | Stop at current stage. Present partial output with explanation. Write diagnostics noting partial completion. |
+
+---
+
+## Token Budget Management
+
+| Checkpoint | Threshold | Action |
+|------------|-----------|--------|
+| After Stage 0 interview | >20% consumed | Warn: "Significant context used in interview. Pipeline will consume more." |
+| After Stage 1 (parallel) | >45% consumed | Warn: "Approaching mid-budget. Consider running Stage 3 sequentially." |
+| After Stage 2 | >60% consumed | Warn: "Context budget constrained. Spawning Stage 3 sequentially." |
+| After Stage 3 | >75% consumed | Warn: "Low budget. Strategist prompt will be minimal." |
+
+---
+
+## Completion Checklist
+
+**IMPORTANT**: Before returning to user, verify ALL items are complete:
+
+- [ ] Stage 0: Arguments parsed (description or --doc)
+- [ ] Stage 0: references/analysis-frameworks.md loaded
+- [ ] Stage 0: templates/final-report.md loaded
+- [ ] Stage 0: AskUserQuestion interview conducted (1-2 rounds)
+- [ ] Stage 0: Synthesized idea brief produced
+- [ ] Stage 1: idea-validator spawned and output verified in `$PROJECT_DIR/logs/`
+- [ ] Stage 1: market-researcher spawned and output verified in `$PROJECT_DIR/logs/`
+- [ ] Stage 2: competitive-analyzer spawned and output verified in `$PROJECT_DIR/logs/`
+- [ ] Stage 3: pattern-documenter spawned and output verified in `$PROJECT_DIR/logs/`
+- [ ] Stage 3: segment-analyzer spawned and output verified in `$PROJECT_DIR/logs/`
+- [ ] Stage 4: strategist spawned and output verified in `$PROJECT_DIR/logs/`
+- [ ] Stage 4: Final report contents read from `$PROJECT_DIR/logs/` and presented to user
+- [ ] Stage 5: Diagnostic YAML written to `$PROJECT_DIR/logs/diagnostics/`

package/skills/product-ideation/references/analysis-frameworks.md

@@ -0,0 +1,161 @@
+# Analysis Frameworks Reference
+
+This file provides the analytical frameworks used by product-ideation sub-agents. Each agent's prompt specifies which sections to apply.
+
+---
+
+## 1. PESTLE Analysis
+
+Use for: market-researcher, competitive-analyzer
+
+PESTLE identifies macro-environmental forces that shape market conditions.
+
+| Factor | Questions to Answer |
+|--------|---------------------|
+| **Political** | Relevant regulations, government policy direction, trade restrictions, stability of operating regions |
+| **Economic** | Market size, GDP trends, consumer spending power, inflation impact, funding climate (VC activity) |
+| **Social** | Demographics of target users, cultural adoption patterns, behavioral shifts, trust in the category |
+| **Technological** | Enabling technologies (AI, APIs, platforms), infrastructure maturity, technology adoption curve |
+| **Legal** | Data privacy laws (GDPR, CCPA), IP landscape, licensing requirements, liability exposure |
+| **Environmental** | Sustainability expectations, supply chain concerns, carbon footprint considerations |
+
+**Application**: Flag any PESTLE factor that is a significant enabler or blocker. Enablers support a BUY signal; blockers shift toward HOLD or SELL.
+
+---
+
+## 2. Porter's Five Forces
+
+Use for: competitive-analyzer
+
+Porter's Five Forces assesses the structural attractiveness of an industry.
+
+| Force | What to Assess | Signal |
+|-------|---------------|--------|
+| **Threat of New Entrants** | Capital requirements, brand moat, switching costs, regulatory barriers | High threat = harder to win |
+| **Bargaining Power of Suppliers** | Dependency on key vendors (cloud, APIs, data), concentration of supply | High power = margin pressure |
+| **Bargaining Power of Buyers** | Customer concentration, price sensitivity, ease of switching | High power = pricing difficulty |
+| **Threat of Substitutes** | Adjacent solutions, DIY alternatives, workarounds | High threat = limited pricing |
+| **Competitive Rivalry** | Number of direct competitors, differentiation, growth rate of market | High rivalry = costly to win |
+
+**Scoring**: Rate each force as Low / Medium / High. Aggregate: 0-1 High = attractive, 2-3 High = mixed, 4-5 High = unattractive.
+
+---
+
+## 3. TAM / SAM / SOM Estimation
+
+Use for: market-researcher, segment-analyzer
+
+| Term | Definition | How to Estimate |
+|------|-----------|-----------------|
+| **TAM** (Total Addressable Market) | Everyone who could theoretically use this type of product | Industry reports, population × usage rate, or comparable market data |
+| **SAM** (Serviceable Addressable Market) | Subset reachable with your specific model, geography, and segment | TAM × (target geography %) × (target segment %) |
+| **SOM** (Serviceable Obtainable Market) | Realistic capture in years 1-3 given competition and go-to-market | SAM × (realistic market share %) — typically 0.5-5% for new entrants |
+
+**Guidance:**
+- Use top-down (industry report → narrow) AND bottom-up (unit count × price) where possible
+- Cite sources with recency (prefer 2023+)
+- If TAM < $500M, flag as niche — consider whether that is a strength (defensible) or weakness (ceiling)
+- If TAM > $50B, flag market crowding risk — large markets attract well-funded competitors
+
+---
+
+## 4. SWOT Analysis Template
+
+Use for: idea-validator, strategist
+
+| Quadrant | Internal | External |
+|----------|----------|----------|
+| **Positive** | Strengths (what the idea has going for it) | Opportunities (market conditions that favor it) |
+| **Negative** | Weaknesses (gaps, risks, resource needs) | Threats (competitors, regulations, trends working against it) |
+
+**Application**: The strategist uses SWOT as a synthesis lens to validate the BUY/HOLD/SELL recommendation. At least 2 items per quadrant.
+
+---
+
+## 5. Jobs-to-be-Done (JTBD) Framework
+
+Use for: segment-analyzer
+
+JTBD focuses on the functional, social, and emotional jobs users hire a product to do.
+
+**Three job types:**
+
+| Type | Definition | Example |
+|------|-----------|---------|
+| **Functional** | The practical task to accomplish | "Track client invoices without spreadsheets" |
+| **Social** | How the user wants to be perceived | "Look professional to clients" |
+| **Emotional** | How the user wants to feel | "Feel in control of my cash flow" |
+
+**Persona structure** (produce at minimum 2 personas):
+
+```
+Persona name: {descriptive name, not a real name}
+Role / context: {who this person is and their situation}
+Primary JTBD (functional): {the core job}
+Primary JTBD (emotional): {the emotional driver}
+Current solution: {what they use today — this is the real competitor}
+Switching trigger: {what would make them switch}
+Willingness to pay: {estimated price range per month or per use}
+Segment size estimate: {rough count within SAM}
+```
+
+---
+
+## 6. BUY / HOLD / SELL Recommendation Thresholds
+
+Use for: strategist
+
+Apply these criteria as a weighted scoring guide, not a rigid checklist. Judgment is required.
+
+### BUY — Strong Opportunity
+
+All three gates must clear, plus at least 4 of the 6 amplifiers:
+
+**Gates (all required):**
+- Market size: TAM >= $1B OR SAM >= $100M with defensible niche
+- Competitive gap: Identifiable unmet need not served by top 3 competitors
+- Timing: Technology or behavioral shift actively enabling this category now
+
+**Amplifiers (4 of 6):**
+- Idea validation verdict: PASS
+- Porter's Five Forces aggregate: 0-2 High forces (attractive industry)
+- PESTLE blockers: 0-1 significant blocker
+- Willingness to pay: At least 1 segment with clear WTP above likely cost basis
+- Pattern evidence: At least 2 documented success patterns apply to this idea
+- Failed competitor lesson: Primary failure reasons are avoidable by this approach
+
+### HOLD — Needs Refinement
+
+Use HOLD when there is genuine merit but material concerns exist:
+
+- Idea validation: CONDITIONAL (merit present but concerns noted)
+- Market exists but gap is not clearly differentiated — needs pivot or sharper focus
+- 2-3 significant PESTLE blockers that are addressable with more planning
+- Timing is early (technology not yet mature) or late (market is consolidating)
+- TAM is attractive but top competitors are very well-funded (>$100M raised)
+
+HOLD output must include: what specific changes would shift this to BUY, and a recommended re-evaluation horizon (e.g., "revisit in 6 months when X matures").
+
+### SELL — Pass on This Idea
+
+Recommend SELL when evidence shows the idea is not viable in its current form:
+
+- Idea validation: FAIL (fundamental feasibility or uniqueness problem)
+- Market is saturated: 5+ funded competitors with no clear differentiation path
+- PESTLE: 2+ insurmountable blockers (e.g., pending regulation that bans the model)
+- TAM < $200M AND no evidence of a defensible niche
+- Failed competitor graveyard: Many attempts with similar approaches have all failed for structural reasons
+
+SELL output must include: primary reasons and, where possible, alternative opportunity areas adjacent to this space.
+
+---
+
+## 7. Confidence Level Definitions
+
+Use for: strategist
+
+| Level | Criteria |
+|-------|----------|
+| **High** | Market data is current and quantified; 5+ competitors profiled; 2+ validated user segments; all pipeline agents completed successfully |
+| **Medium** | Some data gaps; market size estimated rather than sourced; fewer than 5 competitors; 1 validated segment |
+| **Low** | Significant data gaps; market size speculative; competitive landscape thin; 1 or more pipeline stages failed or returned partial output |

package/skills/session-handoff/SKILL.md

@@ -0,0 +1,139 @@
+---
+name: session-handoff
+description: Create consistent session handoff documents for context transfer between sessions. Use when closing a session, reaching 75% token consumption, or when user requests handoff. Ensures proper YAML headers for metrics collection, LF line endings, and complete documentation of progress, decisions, and next steps.
+user-invocable: true
+---
+
+# Session Handoff
+
+Create handoff documents that enable seamless context transfer between sessions.
+
+## Critical Requirements
+
+### Line Endings
+
+Use LF (Unix) line endings only. Never CRLF.
+- Use the Write tool directly (handles line endings correctly)
+- Never copy-paste from Windows clipboard
+- If you see `^M` characters, rewrite the file
+
+### File Naming
+
+**Pattern**: `sessions/session_{N}_{YYYYMMDD}.md`
+
+- `{N}`: Session number (integer, no leading zeros)
+- `{YYYYMMDD}`: Date without separators
+
+Examples: `session_5_20260104.md` (correct), `session_05_20260104.md` (wrong)
+
+## Handoff Template
+
+````markdown
+# Session {N} Handoff
+
+```yaml
+session: {N}
+date: {YYYY-MM-DD}
+phase: "P{X} - {Phase Name}"
+task: "P{X}.{Y} - {Task Name}"
+status: {completed | in_progress | blocked}
+tokens_end: "~{X}K ({Y}%)"
+```
+
+---
+
+## Session Summary
+
+{2-3 sentences on outcomes, not process. What was achieved?}
+
+## What Was Accomplished
+
+- [x] {Completed item with file path}
+- [x] {Completed item with file path}
+- [ ] {Incomplete item - carried forward}
+
+## Files Created/Modified
+
+| File | Action | Lines | Purpose |
+|------|--------|-------|---------|
+| {path} | Created | ~{N} | {Brief description} |
+| {path} | Modified | +{N}/-{M} | {What changed} |
+
+## Verification Status
+
+| Check | Status | Notes |
+|-------|--------|-------|
+| Typecheck | {Pass/Fail/Skipped} | |
+| Lint | {Pass/Fail/Skipped} | |
+| Tests | {Pass/Fail/Skipped} | {X/Y if applicable} |
+
+## Technical Decisions
+
+### {Decision Title}
+- **Decision**: {What was decided}
+- **Rationale**: {Why}
+- **Impact**: {What it affects}
+
+## What's Next
+
+1. {Specific actionable step}
+2. {Next step}
+
+## Blockers / Issues
+
+{List blockers or "None"}
+
+## Learnings
+
+{Patterns discovered or "None"}
+````
+
+## Section Guidelines
+
+### YAML Header (Required)
+
+The YAML header enables metrics collection. All fields required:
+
+| Field | Format | Example |
+|-------|--------|---------|
+| session | Integer | `session: 5` |
+| date | YYYY-MM-DD | `date: 2026-01-04` |
+| phase | Quoted string | `phase: "P1 - Foundation"` |
+| task | Quoted string | `task: "P1.2 - Test Auditor"` |
+| status | Enum | `status: completed` |
+| tokens_end | Quoted string | `tokens_end: "~95K (48%)"` |
+
+### Session Summary
+
+Focus on **outcomes** not process. What was delivered?
+
+Good: "Completed session-handoff skill with CRLF handling. Ready for use."
+Bad: "I started by reading files, then thought about structure, then wrote..."
+
+### What Was Accomplished
+
+- Use `[x]` for complete, `[ ]` for incomplete
+- Include file paths
+- Be specific
+
+### Technical Decisions
+
+Document decisions that affect future work. Include rationale so future sessions understand WHY. Skip if no significant decisions.
+
+### Blockers / Learnings
+
+Always include these sections. Write "None" if empty - don't omit.
+
+## Quick Checklist
+
+Before finalizing:
+
+- [ ] File named `session_{N}_{YYYYMMDD}.md`
+- [ ] YAML header complete and valid
+- [ ] All sections present (even if "None")
+- [ ] Next steps are specific and actionable
+- [ ] No CRLF line endings
+
+## Examples
+
+For detailed examples of completed and in-progress handoffs, see [references/examples.md](references/examples.md).