vkc-visa-assessment-engine

LEE-SANG-BOK's avatarfrom LEE-SANG-BOK

Design and implement the Viet K-Connect visa assessment engine (DB-driven ruleset JSON schema + versioning + effective dates). No hardcoded rules in code. Use for building /api/visa/assess and admin ruleset management.

⭐0stars🔀0forks📁View on GitHub🕐Updated Jan 4, 2026
[Architecture Design]

When & Why to Use This Skill

The VKC Visa Assessment Engine is a specialized architectural framework designed to build scalable, data-driven visa eligibility systems. It replaces hardcoded logic with a flexible DB-driven ruleset, utilizing JSON schema validation, versioning, and effective dating to ensure high maintainability and regulatory compliance for immigration platforms.

Use Cases

  • Automated Eligibility Scoring: Implementing a high-scale API (/api/visa/assess) that calculates the probability of visa approval based on dynamic user data.
  • Dynamic Regulatory Management: Enabling administrators to update complex immigration rules and scoring weights via a database-driven workflow without requiring code redeployments.
  • Audit-Ready Versioning: Managing multiple versions of assessment rules with specific effective dates to ensure historical accuracy and compliance during visa transitions.
  • System Integration: Providing a standardized JSON schema and data model for integrating visa assessment logic into larger immigration or legal-tech platforms.
namevkc-visa-assessment-engine
descriptionDesign and implement the Viet K-Connect visa assessment engine (DB-driven ruleset JSON schema + versioning + effective dates). No hardcoded rules in code. Use for building /api/visa/assess and admin ruleset management.
short-descriptionVisa assessment engine (DB-driven)

VKC Visa Assessment Engine (P1)

Goal

Provide “비자변경 가능성(%)” outputs at scale (cover target-customer visas) while keeping maintenance data-driven:

  • rules/weights live in DB (versioned + effective dates)
  • code is a stable evaluator + storage + admin activation workflow

Non‑negotiable

  • No hardcoding rules in TypeScript. Rules belong to DB rows as JSON.

Core data model (minimum)

  • visa_catalog: visa codes / labels / categories
  • visa_transition_rules: fromVisa→toVisa ruleset JSON + version + effectiveFrom + status(pending|active)
  • visa_assessment_models: score→percent mapping / weights JSON + version + status
  • visa_assessments: user assessment history (probabilityPercent, grade, missing, risks, createdAt)

Required interfaces

  • Ruleset JSON schema (validate before activation)
    • .codex/skills/vkc-visa-assessment-engine/references/ruleset-schema.json
  • API response schema (persist + return)
    • .codex/skills/vkc-visa-assessment-engine/references/response-schema.json
  • User:
    • POST /api/visa/assess (auth + 1/day + save visa_assessments)
  • Admin:
    • Manage visa_catalog
    • Create/update visa_transition_rules (pending)
    • Activate a ruleset version (switch to active)
    • Manage visa_assessment_models version and activation

STEP3 visa registry (SoT)

  • “전체 커버” visa code registry lives in: docs/STEP3_SOT_RESOURCES.md
  • Implementation rule: registry is complete first → rulesets are filled in iteratively (v1→vN)