name: Career Gap Architect description: Analyze the skill gaps between a user’s **CV/Resume** and a target **Job Description (JD)**, then generate a **personalized learning roadmap**, ensuring semantic accuracy for both **Vietnamese** and **English**. allowed-tools: Read, Write, Edit, Glob, Grep

This skill is designed for stable operation in a production environment and is capable of:

•Reading multiple file formats
•Performing multilingual semantic comparison
•Validating AI-generated data
•Optimizing costs through a caching mechanism
•Preventing UI crashes when AI returns malformed output

2. Input Specification

2.1 Primary Inputs

Field	Type	Required	Description
`cv_input`	File \| String	✅	Uploaded CV (PDF, DOCX, DOC) or raw CV text
`jd_input`	File \| String	✅	Uploaded JD or raw job description text
`language_hint`	Enum	❌	`vi`, `en`, `auto` (default: `auto`)
`target_role`	String	❌	Target job role

3. Core Capabilities (Skills)

3.1 File Reader Skill (Multi-format File Parsing)

Functionality

•
Extract text from:
- •.pdf
- •.docx
- •.doc
- •.txt

Standardized Output

json

{
  "raw_text": "string",
  "detected_language": "vi | en | mixed",
  "metadata": {
    "file_type": "pdf | docx | doc | text",
    "page_count": number
  }
}