Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Une commande directe contourne le prompt de vérification. Examinez la source avant de l'exécuter.
Orchestrate all research and discovery activities for product/startup development. This domain skill combines idea discovery, product research, and project naming into a unified research workflow.
Note: Full methodology details available in methodologies/ folder.
Section 1: Idea Discovery
Workflow
1. Gather Context (AskUserQuestion)
- Skills, interests, resources
↓
2. Generate Ideas (faion-research-agent mode: ideas)
- Apply 7 Ps framework + other methods
↓
3. User Selection (AskUserQuestion)
- Pick 3-5 ideas to research
↓
4. Pain Point Research (faion-research-agent mode: pains)
- Reddit, forums, reviews mining
↓
5. Niche Evaluation (faion-research-agent mode: niche)
- Market size, competition, barriers
↓
6. Present Results
↓
User selects idea? → YES → Write to product_docs/idea-validation.md
→ NO → Loop back to step 2
Phase 1: Gather Context
Use AskUserQuestion:
Question 1: "What are your skills/experience?"
Options:
- Software development
- Design/UX
- Marketing/Sales
- Domain expertise (specific industry)
Question 2: "What motivates you?"
Options:
- Solve my own problem
- Big market opportunity
- Passion project
- Side income
Question 3: "How much time can you invest?"
Options:
- Nights & weekends (side project)
- Part-time (20h/week)
- Full-time
Phase 2: Generate Ideas
Call faion-research-agent (mode: ideas) agent:
Task(
subagent_type="faion-research-agent (mode: ideas)",
prompt=f"""Generate startup ideas for:
Skills: {skills}
Motivation: {motivation}
Time commitment: {time}
Use frameworks: 7 Ps, Paul Graham, Personal Pain Points
Generate 15-20 ideas across categories.
""",
description="Generate startup ideas"
)
Phase 3: User Selection
Present ideas grouped by category. Use AskUserQuestion:
Question: "Which ideas do you want to research deeper?"
multiSelect: true
Options:
- {idea1} - {category}
- {idea2} - {category}
- {idea3} - {category}
- {idea4} - {category}
- "None, generate different ones"
Phase 4: Pain Point Research
For each selected idea, call faion-research-agent (mode: pains):
For ideas with validated pain points, call faion-research-agent (mode: niche):
Task(
subagent_type="faion-research-agent (mode: niche)",
prompt=f"""Evaluate niche for: {idea}
Criteria:
- Market size (TAM/SAM/SOM estimates)
- Competition level (red/blue ocean)
- Entry barriers (technical, regulatory, capital)
- Profitability potential (unit economics)
- Your fit (skills match)
Score each 1-10, provide total score.
""",
description="Evaluate niche viability"
)
Phase 6: Present Results & Decision
Show evaluation results. Use AskUserQuestion:
Question: "Evaluation results. What's next?"
Options:
- {idea1} (score: 42/50) - select for development
- {idea2} (score: 38/50) - select for development
- Generate new ideas
- Research other niches for {ideaX}
Phase 7: Write to Documentation
When user selects idea, create aidocs/sdd/{project}/product_docs/idea-validation.md:
Entrepreneurs don't know which features are missing in the market.
Framework
Analysis Process:
Feature Inventory
List all features across top 5 competitors
Categorize: Core, Differentiator, Nice-to-have
Feature Matrix
Competitors as columns
Features as rows
Mark: Has (Y), Partial (P), Missing (N)
Gap Identification
Features no one has
Features only 1-2 have (opportunity)
Features everyone has (table stakes)
Gap Validation
Is this gap intentional (hard, unprofitable)?
Do customers want it? (check reviews)
Can we build it better?
Templates
Feature Matrix:
| Feature | Us | Comp A | Comp B | Comp C | Gap? |
|---------|-------|--------|--------|--------|------|
| Core 1 | Y | Y | Y | Y | No (table stakes) |
| Core 2 | Y | Y | P | Y | No |
| Diff 1 | Y | N | N | P | Yes - Opportunity |
| Diff 2 | P | Y | Y | N | Build out |
| Nice 1 | N | N | N | N | Validate demand |
Gap Validation:
## Gap: {Feature Name}### Evidence- Reviews mentioning need: X
- Forum discussions: X links
- Search volume: X/month
### Why Competitors Don't Have It- {reason}: Technical difficulty / Low priority / Not aware
### Our Advantage- {why we can build it}
### Recommendation- Pursue / Investigate / Skip
Examples
Email Marketing Tools Gap:
Gap: Native A/B testing for subject lines
Evidence: 50+ feature requests in Mailchimp community
Why missing: Requires ML infrastructure
Our advantage: Have ML expertise
Agent
faion-research-agent (mode: competitors)
M-RES-009: User Interviews
Problem
Entrepreneurs build based on assumptions rather than validated user needs.
Framework
Interview Types:
Type
When
Goal
Questions
Problem
Early
Validate problem exists
Open-ended exploration
Solution
After problem validated
Test solution fit
Prototype feedback
Usability
With prototype
Test UX
Task-based
Interview Structure (45 min):
Warm-up (5 min): Build rapport
Current state (10 min): How do they do it today?
Pain exploration (15 min): What's frustrating?
Solution probing (10 min): Would X help?
Wrap-up (5 min): Would they pay? Referrals?
Golden Rules:
Talk less than 20% of the time
Never pitch, only probe
Ask about past behavior, not future intent
Look for emotion (frustration, excitement)
Sample Questions:
"Tell me about the last time you [problem]..."
"What did you try? What happened?"
"On a scale of 1-10, how frustrating is this?"
"How much time/money does this cost you?"
Templates
Interview Script:
## Warm-up
"Thanks for joining. I'm researching [topic]. No sales pitch, just learning."
## Current State
"Walk me through how you currently [task]."
"What tools do you use?"
"How often do you do this?"
## Pain Exploration
"What's the most frustrating part?"
"Tell me about the last time it went wrong."
"What workarounds have you tried?"
## Solution Probing (show concept/prototype)
"Would something like this help?"
"What's missing?"
"Would you pay for this? How much?"
## Wrap-up
"Who else should I talk to?"
"Can I follow up in 2 weeks?"
Examples
Key Insights from 10 Interviews:
8/10 mentioned same frustration
Average time wasted: 3 hours/week
Willingness to pay: $50-100/month
Quote: "I would kill for this solution"
Agent
faion-research-agent (mode: validate)
M-RES-010: Jobs To Be Done (JTBD)
Problem
Features don't connect to real user motivations.
Framework
JTBD Statement:
When I [situation/trigger],
I want to [motivation/action],
So I can [expected outcome/benefit].
Job Dimensions:
Dimension
Question
Example
Functional
What task?
Send invoice
Emotional
How feel?
Confident, professional
Social
How perceived?
Reliable, organized
Job Mapping Process:
Identify situation triggers
Map functional jobs
Uncover emotional jobs (harder)
Discover social jobs (hardest)
Prioritize by frequency + importance
Hiring/Firing Framework:
Why do customers "hire" solutions?
Why do they "fire" (abandon) solutions?
What makes them switch?
Templates
JTBD Canvas:
## Job: {Name}### Situation/Trigger
When I {situation}...
### Functional Job
I want to {action}...
### Emotional Job
So I can feel {emotion}...
### Social Job
And be seen as {perception}...
### Current Solutions- {solution 1}: Hired because {reason}, fired because {reason}
- {solution 2}: Hired because {reason}, fired because {reason}
### Our Opportunity
We can {do X better} by {approach}
Examples
Invoicing Software JTBD:
Situation: When I finish a project for a client
Functional: I want to send a professional invoice quickly
Emotional: So I can feel confident I'll get paid
Social: And be seen as a legitimate business
Agent
faion-research-agent (mode: personas)
M-RES-011: User Persona Creation
Problem
Teams build for abstract "users" rather than specific people with distinct needs.
Framework
Persona Components:
Component
Description
Source
Demographics
Age, role, income, location
Analytics, surveys
Behaviors
Tools used, habits, workflows
Interviews, observation
Pain Points
Top 3-5 frustrations
Interviews, support tickets
Goals
What success looks like
Interviews
Quote
Representative statement
Verbatim from interview
Persona Types:
Type
Priority
Focus
Primary
1
Design for this person
Secondary
2
Consider their needs
Negative
Exclude
Explicitly not for them
Creation Process:
Conduct 5-10 user interviews
Identify patterns (cluster similar users)
Create 2-3 personas (not more)
Validate with team
Post visibly (always reference)
Templates
Persona Template:
# Persona: {Name} ({Type})## Photo
[Stock photo representing this person]
## Demographics-**Age:** 32
-**Role:** Product Manager at startup
-**Income:** $120K
-**Location:** San Francisco
-**Tech savviness:** High
## Behaviors- Works 50+ hours/week
- Uses: Slack, Notion, Figma, Jira
- Checks phone first thing in morning
- Reads newsletters over coffee
## Pain Points1. Too many tools, context switching
2. Stakeholder alignment meetings
3. Data scattered across systems
## Goals- Ship features faster
- Better work-life balance
- Get promoted to Director
## Quote
"I spend more time coordinating than creating."
## Jobs To Be Done- When I [situation], I want to [action], so I can [outcome]
Examples
SaaS Personas:
Primary: "Startup Sarah" - PM at early-stage startup
Secondary: "Enterprise Eric" - PM at Fortune 500
Negative: "Agency Alex" - Freelance consultant
Agent
faion-research-agent (mode: personas)
M-RES-012: Problem Validation
Problem
Entrepreneurs build solutions for problems that don't exist or aren't painful enough.
Framework
Validation Criteria:
Criterion
Threshold
How to Measure
Frequency
Weekly+
"How often do you face this?"
Intensity
7+/10
"How painful is this? (1-10)"
Willingness to Pay
Yes
"Would you pay to solve this?"
Search Behavior
Exists
Check search volume
Competition
Exists
Someone trying to solve it
Evidence Types:
Type
Strength
Source
Verbatim quotes
Strong
Interviews
Forum discussions
Medium
Reddit, communities
Review complaints
Medium
App stores, G2
Search volume
Medium
Google Trends, Ahrefs
Competitor existence
Weak
Market research
Validation Process:
State the problem hypothesis
Define evidence needed
Collect evidence (10+ data points)
Assess against criteria
Decide: Proceed / Pivot / Kill
Templates
Problem Validation Report:
## Problem: {Statement}### Hypothesis
{Who} struggles with {what} because {why}
### Evidence Collected
| Type | Source | Finding |
|------|--------|---------|
| Interview | User 1 | "I spend 3 hours/week on this" |
| Interview | User 2 | "Would pay $50/month" |
| Forum | Reddit | 50 upvotes on complaint post |
| Review | G2 | "Missing feature X" (repeated 10x) |
| Search | Google | "solve X problem" - 5K/month |
### Assessment
| Criterion | Score | Evidence |
|-----------|-------|----------|
| Frequency | Weekly | Interviews |
| Intensity | 8/10 | Interviews |
| WTP | Yes | 4/5 would pay |
| Search | Medium | 5K/month |
| Competition | Yes | 3 competitors |
### Decision**PROCEED** - Strong problem-solution fit
Examples
Validated Problem:
Hypothesis: Freelancers struggle tracking time across projects
Evidence: 8/10 interviews confirmed, avg 5 hrs/week lost
Decision: Proceed to solution validation
Agent
faion-research-agent (mode: validate)
M-RES-013: Pain Point Mining
Problem
Entrepreneurs don't know where customers express frustrations.
Framework
Mining Sources:
Source
Search Strategy
Signal Strength
Reddit
"r/{niche} + frustrating/hate/problem"
High (honest)
Twitter/X
"{product} sucks" OR "wish {product}"
High (real-time)
App Reviews
1-3 star reviews
High (specific)
Forums
Product-specific communities
Medium
Quora
"{problem} solution"
Medium
LinkedIn
Industry discussions
Low (filtered)
Search Queries:
"{keyword} frustrating" site:reddit.com
"{competitor} alternative" site:reddit.com
"{product} review" 1-star
"I wish {tool} could"
"hate using {tool}"
"looking for {solution}"
Analysis Process:
Collect 50+ pain point mentions
Categorize by theme
Count frequency per theme
Note intensity language
Identify gaps in solutions
Templates
Pain Point Mining Report:
## Topic: {Area}### Sources Searched- Reddit: r/x, r/y, r/z
- App Store: {app1}, {app2}
- Forums: {forum1}
### Pain Points Identified
| Theme | Frequency | Sample Quote | Intensity |
|-------|-----------|--------------|-----------|
| Slow sync | 23 mentions | "Takes forever to sync" | High |
| Missing feature X | 15 mentions | "Why doesn't it have X?" | Medium |
| Expensive | 12 mentions | "Not worth $99/month" | High |
### Key Insights1. {insight 1}
2. {insight 2}
### Opportunity
{What solution addresses top pain points}
Examples
Project Management Pain Mining:
Top pain: "Too many clicks to create task" (35 mentions)
Opportunity: One-click task creation from anywhere
Agent
faion-research-agent (mode: pains)
M-RES-014: Niche Viability Scoring
Problem
Entrepreneurs can't objectively assess if a niche is worth pursuing.
Entrepreneurs compete in crowded markets instead of creating new ones.
Framework
Red vs Blue Ocean:
Red Ocean
Blue Ocean
Compete in existing market
Create uncontested space
Beat the competition
Make competition irrelevant
Exploit existing demand
Create new demand
Value-cost trade-off
Break value-cost trade-off
Four Actions Framework:
Action
Question
Result
Eliminate
What factors can we eliminate?
Remove costly/unnecessary
Reduce
What can we reduce below standard?
Simplify
Raise
What can we raise above standard?
Differentiate
Create
What new factors can we create?
Innovate
Strategy Canvas:
X-axis: Key competing factors
Y-axis: Offering level (low to high)
Plot competitors and your new curve
Templates
Blue Ocean Canvas:
## Industry: {Name}### Current Red Ocean Factors
| Factor | Industry Level | Customer Value |
|--------|---------------|----------------|
| Price | High | Medium |
| Features | Many | Low (unused) |
| Support | 24/7 | Low (rarely needed) |
### Four Actions#### Eliminate- {factor}: Why? {reason}
#### Reduce- {factor}: From {X} to {Y}
#### Raise- {factor}: From {X} to {Y}
#### Create- {new factor}: {description}
### New Value Curve
[Strategy canvas visualization]
### Blue Ocean Opportunity
{Description of uncontested space}
Examples
Cirque du Soleil:
Eliminated: Animals, star performers, aisle concessions
Great names are unusable due to domain/handle unavailability.
Framework
Check Priority:
Type
Priority
Importance
.com
1
Essential for credibility
.io
2
Acceptable for tech
.co
3
Alternative
GitHub
1
Essential for open source
Twitter/X
2
Important for marketing
LinkedIn
3
Nice to have
Availability Actions:
Status
Action
Available
Register immediately
Premium ($X)
Consider if <$5K
Taken (parked)
Check price, usually overpriced
Taken (active)
Move to next name
Alternative Strategies:
Add "get", "try", "use" prefix: getTaskFlow.com
Add "app", "hq" suffix: taskflowhq.com
Different TLD: taskflow.io
Creative spelling: taskflw.com
Templates
Domain Check Report:
## Name: {name}### Domain Availability
| Domain | Status | Price | Notes |
|--------|--------|-------|-------|
| {name}.com | Available | $12/yr | Register now |
| {name}.io | Taken | - | Active site |
| {name}.co | Premium | $2,500 | Parked |
### Social Handles
| Platform | Handle | Status |
|----------|--------|--------|
| Twitter | @{name} | Available |
| GitHub | {name} | Taken |
| LinkedIn | /company/{name} | Available |
### Trademark Check- USPTO: No conflicts found
- Note: Not legal advice, consult attorney
### Recommendation**Register {name}.com and @{name} immediately**### Alternatives if Unavailable1. get{name}.com
2. {name}app.com
3. {name}.io
Examples
TaskFlow Availability:
taskflow.com: Taken (active SaaS)
gettaskflow.com: Available
taskflow.io: Available
@taskflow: Taken
@gettaskflow: Available
Recommendation: gettaskflow.com + @gettaskflow
Agent
faion-domain-checker-agent
M-RES-019: Pricing Model Selection
Problem
Entrepreneurs choose wrong pricing models that limit growth or revenue.
Framework
Pricing Models:
Model
Best For
Pros
Cons
Freemium
High volume, low marginal cost
Viral growth
Low conversion (2-5%)
Subscription
Recurring value
Predictable revenue
Churn risk
Usage-based
Variable consumption
Scales with value
Unpredictable revenue
One-time
Complete product
Simple
No recurring revenue
Tiered
Diverse segments
Captures more value
Complex
Per-seat
Team tools
Clear pricing
Resistance to add users
Selection Criteria:
Factor
Question
Value delivery
Continuous or one-time?
Usage patterns
Consistent or variable?
Customer type
Individual or team?
Competition
What do they charge?
Marginal cost
Cost to serve additional user?
Pricing Research:
List top 5 competitors' pricing
Calculate value delivered
Survey willingness to pay (Van Westendorp)
Test with early users
Templates
Pricing Strategy:
## Product: {Name}### Competitor Pricing
| Competitor | Model | Price Range | Notes |
|------------|-------|-------------|-------|
| {name} | Subscription | $X-Y/mo | |
| {name} | Freemium | Free-$X | |
### Value Analysis- Cost of problem: $X/month
- Time saved: X hours/month
- Value captured: 10-20% of savings
### Recommended Model**Tiered Subscription**### Pricing Tiers
| Tier | Price | Features | Target |
|------|-------|----------|--------|
| Free | $0 | {features} | Try before buy |
| Pro | $19/mo | {features} | Individuals |
| Team | $49/mo | {features} | Small teams |
| Enterprise | Custom | {features} | Large orgs |
### Justification- Freemium drives awareness
- Pro captures 80% of revenue
- Enterprise for large accounts
Examples
SaaS Pricing Decision:
Model: Tiered subscription
Free: 1 project, basic features
Pro ($19): Unlimited projects
Team ($49/user): Collaboration features
Justification: Match competitor pricing, capture team value
Agent
faion-research-agent (mode: pricing)
M-RES-020: Customer Interview Framework
Problem
Interviews fail to extract actionable insights due to poor questions and structure.
Framework
Interview Types:
Type
Stage
Goal
Duration
Discovery
Problem validation
Understand context
30 min
Problem
Problem validation
Deep dive on pain
45 min
Solution
Solution validation
Test concepts
45 min
Usability
Product development
Test prototype
60 min
Interview Flow:
1. Warm-up (5 min)
- Build rapport
- Set expectations
↓
2. Context (10 min)
- Their role, background
- Current situation
↓
3. Deep dive (20 min)
- Pain exploration
- Past behavior
- Emotional triggers
↓
4. Concept test (10 min)
- Show solution
- Get reaction
↓
5. Wrap-up (5 min)
- Willingness to pay
- Referrals
Question Types:
Type
Purpose
Example
Open
Explore
"Tell me about..."
Probing
Go deeper
"Why is that?"
Clarifying
Understand
"Can you give an example?"
Summary
Confirm
"So you're saying..."
Anti-patterns:
Leading questions: "Don't you think X is great?"
Future hypotheticals: "Would you use...?"
Pitching: Explaining your solution too early
Multiple questions: Asking 2+ questions at once
Templates
Interview Guide:
## Interview: {Type}### Preparation- Hypothesis to test: {hypothesis}
- Key questions: 5 max
- Recording setup: {yes/no}
### Script#### Warm-up
"Thanks for joining. I'm researching [topic]. No right/wrong answers, just honest feedback."
#### Context
"Tell me about your role."
"Walk me through a typical day."
#### Deep Dive
"Tell me about the last time you [problem]."
"What happened? What did you do?"
"How did that make you feel?"
"What would you change?"
#### Concept Test
[Show prototype/concept]
"What do you think this is?"
"Would this help? How?"
"What's missing?"
#### Wrap-up
"Would you pay for this? How much?"
"Who else should I talk to?"
### Notes Template
| Question | Answer | Insight |
|----------|--------|---------|
| {question} | {verbatim} | {interpretation} |
Examples
Key Interview Insights:
8/10 mentioned same frustration
Verbatim: "I would pay $100 to not deal with this"
Behavior: Currently using 3 tools as workaround
Insight: Strong problem-solution fit
Agent
faion-research-agent (mode: validate)
faion-research-domain-skill v1.1Merged from: faion-idea-discovery, faion-product-research, faion-project-namingMethodologies: M-RES-001 to M-RES-020 (20 total)Agents: 8 research + 2 naming
