Portfolio Composition Skill

Portfolio Composition Skill

Purpose

Transform raw project data into professional case study narratives that demonstrate reasoning, technical decisions, and measurable results.

Output characteristics:

• Reads as professional portfolio entry, not marketing content
• Factual, reflective, confident tone
• Emphasizes process and decision-making over promotional claims
• Uses past tense verbs (designed, validated, implemented)

Avoid:

• Sales language, exclamation marks, generic adjectives
• First-person pronouns unless style requires it
• CTAs, slogans, marketing claims

Required Input

Core fields:

• project_name: Project title
• role: Your role in the project
• timeline: Project duration
• objective: Primary goal
• challenge: Main problem addressed
• methods: Approaches and techniques used
• impact: Results achieved

Optional fields:

• tools: Technologies and tools used
• metrics: Quantified results
• feedback: User or stakeholder quotes
• reflection: Lessons learned
• related_projects: Connected work

Output Structure

Generate JSON with six sections. Each section maximum 120 words.

1. Overview

Summarize what the project is and why it mattered.

Uses: project_name, timeline, role, objective

Format:

{
  "overview": {
    "headline": "string (project title, max 10 words)",
    "summary": "string (2-3 sentences on goal and scope)",
    "meta": { "role": "string", "timeline": "string" }
  }
}

Requirements:

• Neutral tone
• Must mention purpose and scale
• Headline under 10 words

2. Context

Explain environment, target users, and constraints.

Uses: objective, tools, constraints

Format:

{
  "context": {
    "background": "string (setting and target users)",
    "constraints": ["string (technical or business limits, max 10 words each)"],
    "tools_used": ["string"]
  }
}

Requirements:

• Factual descriptions, avoid adjectives like "innovative"
• Each constraint under 10 words

3. Challenge

State the main problem and insights that guided direction.

Uses: challenge, insights

Format:

{
  "challenge": {
    "problem_statement": "string (1-2 sentences)",
    "insights": ["string (3-5 insights or pain points)"]
  }
}

Requirements:

• Mention both what was wrong and why it mattered
• Use active phrasing: "Users struggled to...", "The system lacked..."

4. Process

Detail reasoning, iterations, and methods used.

Uses: methods, decisions, iterations

Format:

{
  "process": [
    {
      "step_title": "string (phase or milestone, max 5 words)",
      "approach": "string (what was done and why, max 40 words)",
      "visual_hint": "string (suggested image or diagram)"
    }
  ]
}

Requirements:

• Use 3-5 steps only
• Focus on how challenges were solved, not listing tools
• Each step_title under 5 words, approach under 40 words

5. Outcome

Show tangible results and evidence of success.

Uses: impact, metrics, feedback

Format:

{
  "outcome": {
    "results": ["string (quantified or descriptive outcomes)"],
    "metrics": ["string (e.g., '+25% retention')"],
    "feedback": ["string (user or stakeholder quotes)"],
    "visual_hint": "string (suggested before/after or result visual)"
  }
}

Requirements:

• Include at least one quantifiable impact or quote
• Avoid adjectives like "amazing" or "great"

6. Reflection

Demonstrate learning and professional growth.

Uses: reflection, next_steps, related_projects

Format:

{
  "reflection": {
    "lessons": ["string (what was learned)"],
    "future_focus": "string (what could be improved or extended)",
    "related_projects": ["string (other works)"]
  }
}

Requirements:

• Authentic and forward-looking tone
• Mention at least one personal or team insight

Validation Checklist

Before finalizing, verify:

• All six sections present (overview through reflection)
• Challenge clearly links to process decisions
• Outcome includes measurable or testimonial proof
• Reflection mentions learning or next iteration
• No sales or CTA language detected
• Each section under 120 words
• Consistent past tense verbs throughout

Example Usage

Input:

{
  "project_name": "Vibe Coding Agent",
  "role": "Product Manager",
  "timeline": "Q3-Q4 2025",
  "objective": "Enable low-code users to build functional web apps with an AI assistant",
  "challenge": "Non-technical creators struggled to debug and publish complex UIs.",
  "methods": ["multi-model orchestration", "skill composition framework", "user co-design workshops"],
  "impact": "Reduced build-to-publish time by 60%, improved satisfaction scores by 2.3x.",
  "metrics": ["60% faster delivery", "2.3x higher usability rating"],
  "feedback": ["'It feels like the AI reads my mind.' - Beta user"],
  "reflection": "Future focus: expanding Skill marketplace and contextual learning."
}

Output:

{
  "overview": {
    "headline": "Vibe Coding Agent",
    "summary": "An AI-assisted builder that lets non-coders design, debug, and deploy web apps end-to-end.",
    "meta": { "role": "Product Manager", "timeline": "Q3-Q4 2025" }
  },
  "context": {
    "background": "Built within YouWare's creation platform for SMB users aiming to build sites visually.",
    "constraints": ["Low-code user base", "Browser-only runtime"],
    "tools_used": ["Figma", "TypeScript", "OpenAI APIs"]
  },
  "challenge": {
    "problem_statement": "Non-technical creators struggled to debug logic and deploy stable UIs.",
    "insights": [
      "Users lost context across design and code steps",
      "Debug process was opaque",
      "Model suggestions were not reusable"
    ]
  },
  "process": [
    {
      "step_title": "Skill Design",
      "approach": "Defined modular Skill units bridging model reasoning and UI actions.",
      "visual_hint": "System diagram showing Skill orchestration"
    },
    {
      "step_title": "Prototype & Test",
      "approach": "Ran co-design sessions to validate flow and refine UI feedback loops.",
      "visual_hint": "User testing screenshots"
    },
    {
      "step_title": "Integration",
      "approach": "Implemented orchestration and validation logic for agent Skill calls.",
      "visual_hint": "Workflow chart of multi-model routing"
    }
  ],
  "outcome": {
    "results": ["Launch-to-publish time reduced by 60%"],
    "metrics": ["2.3x higher usability rating"],
    "feedback": ["'It feels like the AI reads my mind.' - Beta user"],
    "visual_hint": "Before/after interface comparison"
  },
  "reflection": {
    "lessons": ["Clearer Skill definitions improved reasoning consistency"],
    "future_focus": "Extend Skill sharing and adaptive learning features.",
    "related_projects": ["AI Integration Skill"]
  }
}