Skip to content

Repository files navigation

🧠 Shadow Multi-Agent AI System (Coral Protocol)

An autonomous multi-agent AI system built on Coral Protocol, integrating:

  • Shadow Agent (code generation / task handler)
  • Reviewer Agent (evaluation and feedback)
  • Whisper Agent (voice-to-text)
  • Notion Agent (journaling and log sync)
  • Slack Agent (team notifications)
  • Visual wiring via Coral Studio's scenario builder

πŸš€ Quick Start

1. Clone & Configure

git clone https://github.com/sahiixx/Coral-BlackboxAI-Agent.git shadow-agent
cd shadow-agent
cp ../agents-template.env .env  # Use your actual credentials here

2. Run via Docker Compose

docker-compose -f docker-compose-shadow.yml up --build

Studio: http://localhost:3000
Server: http://localhost:8080


🌐 Agents Overview

Agent Role Model
shadow_agent Task/code generator GPT-4.1-mini
reviewer_agent Validates and gives feedback GPT-3.5
whisper_agent Converts audio to text Whisper / OpenAI
notion_agent Logs to Notion DB Notion API
slack_agent Sends messages to Slack channel Slack Bot API

πŸ›  Environment Variables

See agents-template.env

SHADOW_API_KEY=...
WHISPER_API_KEY=...
NOTION_API_KEY=...
SLACK_BOT_TOKEN=...

πŸ“’ Scenario

Use shadow-scenario.json in Coral Studio to visualize:

graph TD
  A[Whisper πŸŽ™οΈ] --> B[Shadow 🧠]
  B --> C[Reviewer βœ…]
  C --> D[Notion πŸ“’]
  C --> E[Slack πŸ“©]
Loading

πŸ“¦ Deployment CI/CD (GitHub Actions)

  1. Create .github/workflows/deploy.yml
  2. Add build-and-push steps
  3. Connect secrets: DOCKERHUB_USERNAME, DOCKERHUB_TOKEN

πŸ“ Directory Layout

.
β”œβ”€β”€ Dockerfile
β”œβ”€β”€ docker-compose-shadow.yml
β”œβ”€β”€ application-shadow.yaml
β”œβ”€β”€ shadow-scenario.json
β”œβ”€β”€ agents-template.env
β”œβ”€β”€ sample-agent-transcript.json

🀝 Credits

Built with ❀️ by @sahiix
Based on Coral Protocol

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

Packages

Used by

Contributors

Languages