Creating AI Skills

Input: "I'm a UX researcher who conducts user interviews"
Output: 
Identity: UX researcher specializing in qualitative user research through structured interviews
Process: 
1. Define research objectives and target user segments
2. Create interview guide with open-ended questions
3. Recruit participants matching criteria
4. Conduct 30-60 minute semi-structured interviews
5. Analyze responses for patterns and insights
6. Synthesize findings into actionable recommendations
Quality Standards: 8+ interviews per research question, documented insights with direct quotes, actionable recommendations tied to business objectives

Progress:

• Parse expertise description for core methodology
• Extract professional role and domain specialization
• Identify specific tools, frameworks, or standards used
• Map methodology into numbered, actionable steps
• Define measurable quality benchmarks
• Create concrete input/output examples
• Rate clarity, specificity, and usability (0-100)

Step-by-step Process

1. Parse Expertise Description

   • Identify professional role/domain
   • Extract specific methodology mentioned
   • Note tools, frameworks, or standards referenced

2. Create Structured Identity

   • Write clear role definition with specialization
   • Specify professional domain and scope
   • Define boundaries of expertise

3. Document Process

   • Break methodology into numbered steps
   • Make each step actionable and specific
   • Include decision points and alternatives where relevant

4. Establish Quality Standards

   • Define what "good" looks like
   • Set measurable criteria
   • Include common success metrics

5. Generate Examples

   • Create realistic input scenarios
   • Show expected output format
   • Include edge cases if necessary

Example 1: Input: "I'm a senior developer who does code reviews focusing on security and performance" Output:

Identity: Senior software engineer specializing in security-focused code reviews and performance optimization
Process:
1. Scan for security anti-patterns (SQL injection, XSS, auth bypasses)
2. Analyze performance bottlenecks using profiling data
3. Review architecture decisions for scalability
4. Check code maintainability and documentation
5. Provide specific remediation recommendations
Quality Standards: Zero critical security vulnerabilities, sub-100ms response times, 90%+ test coverage

Example 2: Input: "I help startups create fundraising pitch decks" Output:

Identity: Startup fundraising advisor specializing in Series A-C pitch deck optimization
Process:
1. Analyze market positioning against competitors
2. Refine value proposition with TAM/SAM data
3. Structure financial projections with unit economics
4. Craft compelling problem-solution narrative
5. Design investor-ready slide flow
Quality Standards: Clear problem-solution fit, defensible $1B+ market size, realistic 3-year projections