Career Full Review

# Career Full Review — Orchestration Prompt

## Instructions

Execute a **complete career package evaluation** producing a unified Career Score /100.

This is the flagship command. It orchestrates all individual rubrics and prompts into a single cohesive assessment.

## Input Required

- Resume (text or file)
- GitHub profile content (see Data Acquisition below)
- List of top 3–5 projects (with descriptions)
- Target role(s)
- LinkedIn profile content (optional — see Data Acquisition below)
- Experience level (Intern / Junior / Mid / Senior)

## Data Acquisition — How to Get Profile Data

> **Critical:** This skill cannot fetch live web content. All data must come from the user. Be explicit about what you need and don't hallucinate content you haven't been given.

### GitHub
Ask the user to provide ONE of the following:
1. **Paste their GitHub profile page content** — bio, pinned repos, contribution graph description
2. **Paste README files** from their top 3–5 repositories
3. **Describe their GitHub** — pinned repos, languages used, stars, documentation quality

If the user hasn't provided GitHub data, **ask for it explicitly** before scoring this section. Do not guess or infer. If the user declines, score GitHub as "Not Assessed" and adjust the Career Score weights accordingly.

### LinkedIn
Ask the user to provide ONE of the following:
1. **Paste their LinkedIn profile content** — headline, about section, experience, skills
2. **Screenshot** of their profile (if supported by the tool)
3. **Describe their LinkedIn** — headline, summary, endorsements, connections estimate

If not provided, mark LinkedIn as "Not Assessed" and use the weights without LinkedIn.

## Orchestration Sequence

### Step 1: Resume Analysis
- Load `rubrics/ai_resume_rubric.md` for scoring criteria (with calibration anchors)
- Load `prompts/ats_review.md` for ATS evaluation
- Load `prompts/keyword_analysis.md` for keyword gap analysis
- Load `knowledge/resume_best_practices.md` for quality benchmarks and canonical rewrite rules
- Load `knowledge/ai_ml_keywords.md` for role-specific keywords
- Reference `examples/good_resume_example.md` for before/after demonstrations
- **Output: ATS Score /100**

### Step 2: GitHub Analysis (requires user-provided data)
- Load `rubrics/github_rubric.md` for scoring criteria
- Load `prompts/github_review.md` for evaluation process
- Reference `examples/weak_github_example.md` for comparison
- **Precondition:** User must have provided GitHub profile content. If not, ask.
- **Output: GitHub Score /100 (or "Not Assessed")**

### Step 3: Project Portfolio Analysis
- Load `rubrics/project_evaluation_rubric.md` for scoring criteria
- Load `prompts/project_review.md` for evaluation process
- Load `knowledge/project_patterns.md` for strong/weak signal detection
- Reference `examples/strong_project_example.md` for benchmarks
- Score each project individually, then compute weighted average
- **Output: Project Score /100**

### Step 4: Interview Readiness Assessment
- Load `rubrics/interview_rubric.md` for scoring criteria
- Load `prompts/interview.md` for question bank
- Load `knowledge/interview_knowledge.md` for topic coverage
- Reference `examples/interview_answer_example.md` for STAR benchmarks
- Assess based on resume evidence + project depth + stated skills
- **Output: Interview Score /100**

### Step 5: Market Fit Assessment
- Load `knowledge/market_intelligence.md` for market demands
- Load `knowledge/ai_ml_keywords.md` for role alignment
- Cross-reference candidate's skills vs market demand
- ⚠️ **Staleness caveat:** Market data has a shelf life. Note the "Last Updated" date from market_intelligence.md. If it's >6 months old, caveat salary and trend claims with "as of [date] — verify against current postings."
- **Output: Market Fit Score /100**

### Step 6: LinkedIn Assessment (if data provided)
- Load `rubrics/linkedin_rubric.md` for scoring criteria
- Load `prompts/linkedin_review.md` for evaluation process
- **Precondition:** User must have provided LinkedIn content. If not, ask. If declined, skip.
- **Output: LinkedIn Score /100 (or "Not Assessed")**

### Step 7: Personalization (if available)
- Load `knowledge/my_profile.md` for candidate context
- Compare current state vs goals
- Identify specific gaps relative to target companies and roles

## Score Calculation