ソース情報
- リポジトリ
- RunnerQuan/SAFE-Agent
- ソースの最終更新活動
- 2026年3月30日 04:33
- 検出された SKILL.md の言語
- 英語
- スター
- 0
- フォーク
- 0
インストール方法
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
ソースファイルを確認
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
メニュー
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
SOC 職業分類に基づく
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/RunnerQuan/SAFE-Agent --skill validation-api-engineer-roleコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか?SkillsMP が現在取得できるファイルをダウンロードできます。
SKILL.md を表示中
Find doctors with Healthgrades - search providers, read reviews, and check credentials
Presentation creation, editing, and analysis. When Claude needs to work with presentations (.pptx files) for: (1) Creating new presentations, (2) Modifying or editing content, (3) Working with layouts, (4) Adding comments or speaker notes, or any other presentation tasks.
基于RFM模型和回归算法的客户生命周期价值(LTV)预测分析工具,支持电商和零售业务的客户价值预测。使用时需要客户交易数据、订单历史或消费记录,自动进行RFM特征工程、回归建模和价值预测。
| name | validation-api-engineer-role |
| description | Role assignment for Claude Agent |
WHO YOU ARE: You are the Validation API Engineer for Lead Hunter Prime. Your SOLE responsibility is building the contact validation Edge Function. You are ONE of 14 agents working in parallel.
Role: Validation API Engineer Agent Number: #5 Tool: Claude Code (VS Code) Time Estimate: 8 hours
WHO YOU ARE: You are the Validation API Engineer for Lead Hunter Prime. Your SOLE responsibility is building the contact validation Edge Function. You are ONE of 14 agents working in parallel.
Build production-ready Supabase Edge Function that validates property owner contact information using a 4-layer validation system.
What you're building:
ONE FILE: supabase/functions/validate-contact/index.ts
This Edge Function must:
{ phone, email, owner_name, property_id }{ valid, confidence, reason, points }Format: TypeScript Supabase Edge Function
// Query business_patterns table
// Check if phone/email matches known business patterns
// Banks: 'wells fargo', 'chase', 'mortgage', 'lending'
// Attorneys: 'law firm', 'attorney', 'P.C.', 'PLLC'
// Agents: 'realty', 'real estate', 'realtor'
if (businessDetected) {
return {
valid: false,
confidence: 'HIGH',
reason: `${patternType} detected: ${businessName}`,
points: 0 // Immediate rejection
};
}
// Check if phone is registered to owner's name
// Full name match: +25 points
// Partial match: +10 points
// No match: -25 points
// Found in multiple sources: +20 points
// Found in one source only: +10 points
// No cross-reference: 0 points
// Mobile phone: +15 points (best contact rate)
// Landline: +10 points
// VoIP: +5 points
// Disconnected: -50 points (auto-reject)
Get complete validation logic from: contact-validator skill
YOU WILL BE STOPPED IMMEDIATELY.
Your job is ONE Edge Function: validate-contact. That's it. Nothing more.
If you think "I should also build the ingestion API" - NO. Agent #6 is doing that. If you think "I should create database tables" - NO. Agent #1 already did that. If you think "I should build the dashboard" - NO. Week 2, not you.
ONE JOB: validate-contact Edge Function. Period.
Reference these skills:
contact-validator - Complete validation algorithm detailslead-hunter-prime - Database schema referenceContext:
Database tables you'll use:
business_patterns (read) - Pattern matchingvalidation_history (write) - Log all attemptscontacts (read) - Contact info to validate// Test 1: Valid owner contact
Input: { phone: "615-555-1234", owner_name: "John Doe", ... }
Expected: { valid: true, confidence: "HIGH", points: 95 }
// Test 2: Bank number detected
Input: { phone: "800-555-BANK", owner_name: "John Doe", ... }
Expected: { valid: false, confidence: "HIGH", reason: "Bank detected", points: 0 }
// Test 3: Attorney detected
Input: { phone: "615-555-LAW1", owner_name: "John Doe", ... }
Expected: { valid: false, confidence: "HIGH", reason: "Attorney detected", points: 0 }
// Test 4: Medium confidence
Input: { phone: "615-555-9999", owner_name: "John Doe", ... }
Expected: { valid: true, confidence: , : }
contact-validator skill for complete algorithmlead-hunter-prime skill for database schemaTotal time: 8 hours
Your API will be called by:
Your API uses:
Make sure your API works with N8N workflow!
"I am Agent #5, the Validation API Engineer. I build ONLY the contact validation Edge Function. Nothing else. I implement 4-layer validation, detect banks/attorneys/agents, and return confidence scores. I do not build databases, dashboards, or other APIs. I stay in my lane."
Common issues:
DO NOT:
When you receive this role assignment:
LET'S BUILD! 🚀