mirror of https://github.com/thecyberlearn/quantum-ai-v3.git synced 2026-08-18 14:52:58 +00:00

Claude bc662e6af0 🎯 Implement dynamic agent configuration system for ultimate scalability

BREAKTHROUGH: Configuration-driven agent management that scales to 100+ agents

## New System Architecture:
- **JSON Configuration Files**: All agents defined in version-controlled JSON
- **Dynamic Loading**: populate_agents.py reads from config files automatically
- **Zero Code Changes**: Add new agents by creating JSON files only
- **Automatic Railway Sync**: Single command ensures database consistency

## File Structure:
- agents/configs/categories/categories.json - All categories
- agents/configs/agents/*.json - Individual agent configurations
- agents/configs/README.md - Complete documentation

## Key Benefits:
✅ **Ultimate Scalability**: Add 1000+ agents without touching code
✅ **Version Control**: All agent definitions tracked in git
✅ **Railway Consistency**: Single command syncs local and production
✅ **Developer Experience**: JSON files are easier than Python commands
✅ **Validation**: Built-in error handling and field validation
✅ **Documentation**: Self-documenting with clear examples

## Migration Path:
- Extracted all 6 existing agents to JSON configurations
- Updated populate_agents.py to load dynamically from configs
- Maintained backward compatibility
- Added comprehensive documentation

## Usage:
```bash
# Add new agent: Create JSON file in agents/configs/agents/
# Deploy to Railway: python manage.py populate_agents
# Result: Agent automatically appears in marketplace
```

This solves the original issue: "if we create another will it show on railway also???"
Answer: YES - just create JSON file and run populate_agents\!

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

2025-08-05 08:38:59 +05:30

3.4 KiB

Raw Blame History

Agent Configuration System

This directory contains JSON configuration files for dynamically creating agents and categories in the Quantum Tasks AI platform.

Directory Structure

agents/configs/
├── categories/
│   └── categories.json          # All agent categories
├── agents/
│   ├── ai-brand-strategist.json        # Direct access agent
│   ├── cybersec-career-navigator.json  # Direct access agent
│   ├── five-whys-analysis.json         # Chat webhook agent
│   ├── job-posting-generator.json      # Form webhook agent
│   ├── pdf-summarizer.json             # File upload webhook agent
│   └── social-ads-generator.json       # Form webhook agent
└── README.md                    # This file

How It Works

Categories are defined in categories/categories.json
Agents are defined in individual JSON files in agents/
Run python manage.py populate_agents to create all agents from configs
Adding new agents is as simple as creating a new JSON file

Adding New Agents

Step 1: Create JSON Configuration File

Create a new file in agents/ directory, e.g., email-writer.json:

{
  "slug": "email-writer",
  "name": "Email Writer",
  "short_description": "AI-powered professional email writing assistant",
  "description": "Generate professional emails for any purpose with AI assistance.",
  "category": "marketing",
  "price": 3.0,
  "agent_type": "form",
  "system_type": "webhook",
  "form_schema": {
    "fields": [
      {
        "name": "email_type",
        "type": "select",
        "label": "Email Type",
        "required": true,
        "options": [
          {"value": "business", "label": "Business Email"},
          {"value": "marketing", "label": "Marketing Email"}
        ]
      }
    ]
  },
  "webhook_url": "http://localhost:5678/webhook/email-writer",
  "access_url_name": "",
  "display_url_name": ""
}

Step 2: Run Population Command

python manage.py populate_agents

Step 3: Agent Appears Automatically

The agent will now appear in the marketplace with the configured settings.

Agent Types

Webhook Agents (N8N Integration)

Set system_type: "webhook"
Include detailed form_schema with fields
Set webhook_url to N8N endpoint
Leave access_url_name and display_url_name empty

Direct Access Agents (External Forms)

Set system_type: "direct_access"
Set form_schema: {"fields": []}
Set webhook_url to external form URL (JotForm, etc.)
Set access_url_name: "agents:direct_access_handler"
Set display_url_name: "agents:direct_access_display"

Field Types for Webhook Agents

text: Single-line text input
textarea: Multi-line text input
select: Dropdown with options array
file: File upload with drag-and-drop
url: URL input with validation
checkbox: Boolean checkbox

Benefits

✅ Scalable: Add 100+ agents without code changes ✅ Version Controlled: All agent definitions in git ✅ Consistent: Ensures local and Railway databases match ✅ Simple: Just create JSON file and run command ✅ Validated: Built-in validation and error handling

Railway Deployment

On Railway, just run:

python manage.py populate_agents

All agents defined in JSON files will be created automatically, ensuring Railway marketplace shows all agents consistently.

3.4 KiB Raw Blame History