Agentic Recruitment Protocol (ARP)

The Agentic Recruitment Protocol (ARP) is an interaction model and open standard for connecting candidates, their AI agents, employers, and recruitment services to complete hiring processes seamlessly.

The specification is currently maintained by JINZ.AI (@jinzaix) and is currently in draft.

• For employers — Reach more qualified candidates. Hire top talent by making your job postings available for discovery and application through AI agents—all while using your existing ATS infrastructure.
• For AI Agents — Embed recruitment into your application. Let your users discover and apply to jobs directly in your application, without being the employer of record.
• For ATS providers — Grow your volume. Process agentic applications by passing secure candidate data between candidates and employers through AI agents.

Learn more at agenticrecruitment.dev.

📦 Repo Structure

<repo-root>/
├── rfcs/
│   └── rfc.*.md
│
├── spec/
│   ├── openapi/
│   │   ├── openapi.arp_rest.yaml
│   │   └── openapi.arp_mcp.yaml
│   │
│   └── json-schema/
│       └── schema.*.json
│
├── examples/
│   ├── rest/
│   └── mcp/
│
├── changelog/
│   └── *.md
│
├── mock_server/
│   └── server.js
│
├── MAINTAINERS.md
├── CONTRIBUTING.md
├── LICENSE
└── README.md

🔗 Quick Links

🛠 Getting Started

ARP has been first implemented as a mock server, providing a development-ready reference implementation for developers and integrators:

• Mock Server: Complete reference implementation with all endpoints
• OpenAPI Specs: Machine-readable API documentation
• JSON Schemas: Comprehensive data model definitions
• Examples: Ready-to-use request/response samples

To start building with ARP:

1. Review this repo's OpenAPI specs and JSON Schemas.
2. Use the mock server for development and testing.
3. Follow the guides provided in the linked documentation.
4. Test using the examples provided in this repo.

Quick Start

# Clone the repository
git clone https://github.com/jinzaix/agentic-recruitment-protocol.git
cd agentic-recruitment-protocol

# Start the mock server
cd mock_server
npm install
npm start
# Server runs on http://localhost:8787

Try REST Examples

# List job postings
curl -s http://localhost:8787/job_postings | jq

# Create an application
curl -s -X POST http://localhost:8787/applications \
  -H 'Content-Type: application/json' \
  -d @../examples/rest/create-application.json | jq

# Create an offer
curl -s -X POST http://localhost:8787/offers \
  -H 'Content-Type: application/json' \
  -d @../examples/rest/create-offer.json | jq

Try MCP Examples

# Search jobs using MCP
curl -s -X POST http://localhost:8787/mcp \
  -H 'Content-Type: application/json' \
  -d @../examples/mcp/searchJobs.json | jq

# Create application using MCP
curl -s -X POST http://localhost:8787/mcp \
  -H 'Content-Type: application/json' \
  -d @../examples/mcp/createApplication.json | jq

📚 Documentation

🚀 Key Features

Dual Protocol Support

• REST API: Traditional HTTP endpoints for web applications
• MCP Procedures: JSON-RPC 2.0 for AI agent integration
• Unified Data Models: Consistent schemas across both protocols

Comprehensive Workflow

• Job Discovery: Search and filter job postings
• Application Management: Submit and track applications
• Interview Scheduling: Book and manage interviews
• Offer Negotiation: Create, accept, decline, or counter offers
• Contract Generation: Automated contract creation
• Consent Management: Granular data sharing permissions

AI Agent Integration

• Intelligent Job Matching: AI-powered candidate-job matching
• Automated Applications: Streamlined application submission
• Smart Scheduling: Intelligent interview coordination
• Credential Verification: Automated credential validation
• Negotiation Support: AI-assisted offer negotiations

Security & Privacy

• Consent Framework: Granular permission system
• Data Minimization: Share only necessary information
• Audit Trail: Complete activity logging
• Secure Transmission: Encrypted data exchange

SEO & Discoverability

• Search Engine Compatible: Built with Schema.org structured data standards
• Search Engine Optimized: Job postings appear in search engine results
• Rich Snippets: Enhanced visibility with structured data markup
• Industry Standard: Aligns with widely adopted Schema.org vocabulary

🏗 Architecture

Core Components

┌─────────────────┐    ┌─────────────────┐    ┌─────────────────┐
│   Candidate     │    │   Employer      │    │  Recruitment    │
│   AI Agent      │◄──►│   AI Agent      │◄──►│   Service       │
└─────────────────┘    └─────────────────┘    └─────────────────┘
         │                       │                       │
         │                       │                       │
         ▼                       ▼                       ▼
┌─────────────────────────────────────────────────────────────────┐
│                    ARP Protocol Layer                           │
│  ┌─────────────┐  ┌─────────────┐  ┌─────────────┐              │
│  │   REST API  │  │   MCP API   │  │   Events    │              │
│  └─────────────┘  └─────────────┘  └─────────────┘              │
└─────────────────────────────────────────────────────────────────┘

Data Flow

1. Discovery: AI agents search for relevant job postings
2. Application: Candidates apply through AI agents with consent
3. Screening: Automated credential verification and matching
4. Interview: Intelligent scheduling and coordination
5. Offer: AI-assisted negotiation and decision making
6. Contract: Automated contract generation and execution

📝 Contributing

We welcome contributions! See CONTRIBUTING.md for:

• Branching model
• Pull request guidelines
• Spec versioning and review process

All changes must include:

• Updated OpenAPI / JSON Schemas
• New or updated examples
• Changelog entry in changelog/unreleased.md

📜 License

Licensed under the Apache 2.0 License. See LICENSE for details.

About

The Agentic Recruitment Protocol (ARP) is an interaction model and open standard for connecting candidates, their AI agents, employers, and recruitment services to complete hiring processes seamlessly. The specification is currently maintained by Kamol Mavlonov (@kamoljan) and JINZ.AI (@jinzaix).

agenticrecruitment.dev

Resources

• Maintainers
• Contributing
• License
• Changelog

Badges