jmsktm/scout

ID8SCOUT - Validation Engine

Purpose

Transform raw ideas into validated (or invalidated) opportunities through systematic research. Every idea deserves rigorous testing before building.

Output: Validation Report with BUILD / PIVOT / KILL verdict

---

When to Use

• User has a new business/product idea to evaluate
• User asks "should I build X?"
• User wants market research on a concept
• User needs competitive analysis
• User wants to understand if there's demand
• Project is in CAPTURED or VALIDATING state

---

Commands

/scout <idea-description>

Run full validation on an idea.

Process:

1. INTAKE - Clarify the idea
2. MARKET - Size the opportunity
3. COMPETITION - Map the landscape
4. COMMUNITY - Mine real signals
5. CALIBRATE - Adjust for solo builder
6. SYNTHESIZE - Render verdict

/scout market <topic>

Run only market analysis.

/scout competitors <product/space>

Run only competitive analysis.

/scout signals <topic>

Run only community signal mining.

---

Validation Process

Phase 1: INTAKE (2-3 minutes)

Clarify the idea before researching.

Questions to answer:

1. What is it? - One sentence description
2. Who is it for? - Specific target user
3. What problem does it solve? - Pain point or desire
4. How is it different? - Why not use existing solutions?
5. What's the hypothesis? - What are we betting on?

If project exists in tracker:

• Load from project card
• Use existing one-liner and context

If new idea:

• Prompt for clarification
• Suggest creating project: /tracker new

---

Phase 2: MARKET (5-10 minutes)

Size the opportunity and understand trends.

Use: mcp__perplexity__search and mcp__perplexity__reason

Research areas:

1. TAM/SAM/SOM

   • Total Addressable Market - Everyone who could use this
   • Serviceable Addressable Market - Those you can reach
   • Serviceable Obtainable Market - Realistic capture

2. Market Trends

   • Is this market growing or shrinking?
   • What's driving the trend?
   • Where in the adoption curve?

3. Timing Assessment

   • Too early? (Market not ready)
   • Too late? (Saturated)
   • Just right? (Growing, not crowded)

Search queries to run:

- "{topic} market size 2024 2025"
- "{topic} industry trends"
- "{topic} growth rate"
- "{problem} solutions market"

Output: Market Analysis section of validation report

---

Phase 3: COMPETITION (5-10 minutes)

Map the competitive landscape and find gaps.

Use: mcp__perplexity__search, mcp__firecrawl__firecrawl_search, mcp__firecrawl__firecrawl_scrape

Research areas:

1. Direct Competitors

   • Who solves the same problem?
   • What do they charge?
   • How do they position?

2. Indirect Competitors

   • Alternative solutions to the problem
   • Adjacent products that could expand

3. Competitor Weaknesses

   • What do users complain about?
   • What's missing?
   • Where are the gaps?

4. Competitive Moats

   • What advantages do incumbents have?
   • Network effects? Data? Brand?

Search queries to run:

- "{topic} tools alternatives"
- "{competitor} reviews complaints"
- "best {solution type} 2024 2025"
- "{competitor} vs" (autocomplete reveals competitors)

Scraping targets:

• Competitor landing pages (positioning, pricing)
• Review sites (G2, Capterra, Product Hunt)
• Comparison articles

Output: Competitive Landscape section of validation report

---

Phase 4: COMMUNITY (5-10 minutes)

Mine real signals from where users talk.

Use: mcp__perplexity__search, mcp__firecrawl__firecrawl_search

Signal sources:

1. Reddit

   • Subreddits where target users hang out
   • Posts asking for solutions
   • Complaints about existing tools

2. YouTube

   • Video comments on related content
   • Tutorial views (demand signal)
   • Creator complaints

3. Twitter/X

   • Threads about the problem
   • Product complaints
   • Feature requests

4. Forums/Communities

   • Indie Hackers
   • Hacker News
   • Niche communities

What to look for:

• Pain intensity (how much do they care?)
• Willingness to pay (are they spending already?)
• Frequency (how often do they have this problem?)
• Workarounds (what janky solutions do they use?)

Search queries to run:

- "site:reddit.com {problem} help"
- "site:reddit.com {existing solution} frustrating"
- "{topic} indie hackers"
- "{problem} twitter thread"

Output: Community Signals section with real quotes

---

Phase 5: CALIBRATE (2-3 minutes)

Adjust all estimates for AI-augmented solo builder reality.

Calibration factors:

| Traditional Assumption | Calibrated Reality | |------------------------|-------------------| | 6-person dev team | 1-2 people + AI tools | | $100K+ budget | $0-10K bootstrap | | 6-12 month timeline | 2-8 weeks to MVP | | Enterprise features first | Core value only | | VC-scale growth | Sustainable indie growth |

Recalibrate:

• Development estimates (AI makes this 5-10x faster)
• Market capture (smaller is fine for solo)
• Competitive moats (speed and focus beat resources)
• Revenue goals ($10K MRR can be life-changing)

Apply to:

• Execution plan timeline
• Budget requirements
• Success metrics

---

Phase 6: SYNTHESIZE (5 minutes)

Combine all research into verdict.

Verdict options:

| Verdict | Meaning | Criteria | |---------|---------|----------| | BUILD | Proceed to architecture | Evidence of demand + achievable by solo builder + acceptable competition | | PIVOT | Refine and revalidate | Core insight valid but approach needs change | | KILL | Don't build this | No demand, or insurmountable competition, or poor fit |

Confidence levels:

• High - Strong evidence, clear signals
• Medium - Mixed signals, some uncertainty
• Low - Weak evidence, significant unknowns

For BUILD verdict, also provide:

• Recommended MVP scope
• Key risks to watch
• Suggested first milestone

For PIVOT verdict, also provide:

• What to change
• New hypothesis to test
• Suggested research focus

For KILL verdict, also provide:

• Why it won't work
• What we learned
• Any salvageable ideas

---

Integration with Tracker

On Start

If project slug provided:
  Load project from tracker
  Use existing context
Else:
  Prompt to create project first
  Or run as standalone research

On Complete

1. Generate validation report
2. Save to project artifacts (if project exists)
3. Log to tracker:
   /tracker log {project} "Scout: Validation complete - {VERDICT} verdict ({confidence})"
4. If BUILD: Suggest /tracker update {project} VALIDATED
5. If KILL: Suggest /tracker kill {project}

---

MCP Usage Patterns

Perplexity (Primary Research)

# Quick market search
mcp__perplexity__search({
  query: "{topic} market size trends 2024 2025"
})

# Complex reasoning
mcp__perplexity__reason({
  query: "Analyze the competitive landscape for {product type}. Who are the major players, what are their strengths and weaknesses, and where are the gaps?"
})

# Deep research
mcp__perplexity__deep_research({
  query: "Comprehensive analysis of {market}",
  focus_areas: ["market size", "growth trends", "key players", "emerging opportunities"]
})

Firecrawl (Web Scraping)

# Search for competitors
mcp__firecrawl__firecrawl_search({
  query: "{product type} tools alternatives",
  limit: 10
})

# Scrape competitor landing page
mcp__firecrawl__firecrawl_scrape({
  url: "https://competitor.com",
  formats: ["markdown"]
})

# Extract structured data
mcp__firecrawl__firecrawl_extract({
  urls: ["https://competitor.com/pricing"],
  prompt: "Extract pricing tiers, features, and target audience",
  schema: {
    type: "object",
    properties: {
      pricing_tiers: { type: "array" },
      features: { type: "array" },
      target_audience: { type: "string" }
    }
  }
})

GitHub (Competitor Analysis)

# Search for open source alternatives
mcp__github__search_repositories({
  query: "{topic} tool"
})

# Check competitor repo activity
mcp__github__list_commits({
  owner: "competitor",
  repo: "their-product"
})

---

Subagent Integration

market-intelligence-analyst

Delegate complex research tasks:

Task({
  subagent_type: "market-intelligence-analyst",
  prompt: "Research the {topic} market. Focus on:
    1. Market size and growth rate
    2. Key trends driving the market
    3. Major players and their positioning
    4. Emerging opportunities and gaps

    Provide specific data points with sources."
})

---

Output Format

Generate validation report using templates/validation-report.md

Key sections:

1. Executive Summary (verdict + confidence + one paragraph)
2. Market Analysis (TAM/SAM/SOM, trends, timing)
3. Competitive Landscape (players, gaps, moats)
4. Community Signals (real quotes, pain intensity)
5. Founder Fit Assessment (skills, resources, motivation)
6. Execution Plan (calibrated for solo builder)
7. Risks & Mitigations
8. Verdict & Reasoning
9. Sources

---

Error Handling

| Issue | Response | |-------|----------| | MCP unavailable | Fall back to web search, note limitation | | Insufficient data | Lower confidence, note gaps | | Conflicting signals | Present both sides, explain uncertainty | | No competitors found | Flag as potential blue ocean OR hidden problem | | Overwhelming competition | Analyze for niche opportunities |

---

Changelog

v1.0.0 (2025-12-21)

• Initial release
• Full validation pipeline
• Perplexity + Firecrawl integration
• Calibration layer for solo builders
• Tracker integration