Zero‑Exposure Slack Sender (MGC Secure Edition) MCP Server

Zero‑Exposure Slack Sender (MGC Secure Edition)

Send Slack messages safely without exposing webhook URLs or bot tokens.
All credentials are stored locally in MGC Blackbox, ensuring zero plaintext secrets and zero accidental leaks.

---

✨ Features

• Zero‑Exposure: Slack credentials never appear in code or logs
• Local encryption: Powered by MGC Blackbox secure vault
• Supports Webhook & Bot Token
• CLI + MCP Tool dual usage
• Mock mode for development without real Slack

---

📦 Installation

Install MGC Blackbox:

pip install mgc-blackbox
mgc

Clone this skill:

git clone https://github.com/zkeviny/MGC-Blackbox
cd MGC-Blackbox/mgc_skill/secure_slack_bot

Run the MCP server:

python mcp_server.py

---

🔐 Store Slack Credentials in MGC

Create a JSON file:

{
  "webhook_url": "https://hooks.slack.com/services/T.../B.../xxxx",
  "bot_token": "xoxb-...",
  "default_channel": "#general"
}

Store it securely:

mgc_save info_type=config info_owner=slack_bot < slack_config.json

info_type and info_owner are identifiers you define yourself, used to retrieve encrypted data from MGC.

---

🚀 Usage

▶ Command Line

# Basic message
python example.py -t "Deployment complete!"

# Specify channel
python example.py -t "Deploy done" -c "#deployments"

# Custom username + icon
python example.py -t "Deploy done" -u "Deploy Bot" -i ":rocket:"

# Mock test (no real Slack request)
python example.py -t "Test" --mock

---

🛠 MCP Tool: send_slack_message

Example:

from mcp import MCPClient

client = MCPClient("http://localhost:57219")

result = client.call_tool("send_slack_message", {
    "text": "Deployment complete!",
    "channel": "#deployments",
    "username": "Deploy Bot",
    "icon_emoji": ":rocket:"
})

print(result)

Tool parameters:

Name	Type	Description
text	string	Message text
channel	string	Slack channel (optional)
username	string	Custom bot name (optional)
icon_emoji	string	Emoji icon (optional)

---

📁 Files

File	Description
README.md	This file
SKILL.md	Skill documentation for AI clients
mcp_server.py	MCP server implementation
example.py	CLI demo script

---

🔒 Security

• Credentials stored only in local MGC Blackbox
• No plaintext secrets in code or logs
• No cloud upload
• No subprocess / exec / shell usage
• Safe for AI automation workflows

---

📄 License

MIT
Author: MirginCipher

---