📧 Daily Standup Email Agent

An AI-powered agent that collects daily standup updates, structures them into a professional Minutes of Meeting (MOM) email, and sends it via Zoho Mail MCP. Built with LangChain, LangGraph, and Groq's blazing-fast LLM inference (Llama 3.3 70B).

⚡ How It Works

1. You provide standup updates (text or interactive)
2. The agent structures them into a professional MOM email using Groq AI
3. Zoho Mail MCP sends the email through your Zoho Mail account

No SMTP credentials needed in the agent — email delivery is handled entirely by the Zoho Mail MCP server configured in your IDE.

🏗️ Architecture

┌──────────────┐
│ Parse Input   │  ← Extract members, tasks, recipients
└──────┬───────┘
       │
┌──────┴───────┐
│ Compose Email │  ← Generate plain text + HTML body
└──────┬───────┘
       │
┌──────┴───────┐
│ Subject Line  │  ← Generate professional subject
└──────┬───────┘
       │
┌──────┴───────┐
│ Prepare MCP   │  ← Build Zoho Mail MCP payload
└──────────────┘
       │
       ▼
┌──────────────────┐
│ Zoho Mail MCP    │  ← IDE sends via MCP (external)
│ (send_mail tool) │
└──────────────────┘

📦 Installation

cd daily-standup-agent
npm install
npm link

🔧 Setup

1. Agent Setup (Groq + Recipients)

daily-standup setup

2. Zoho Mail MCP Setup (in IDE)

Configure the Zoho Mail MCP server in your IDE's MCP config (.kiro/settings/mcp.json):

1. Go to mcp.zoho.com
2. Create a new MCP server
3. Add the Zoho Mail tools (specifically send_mail)
4. Copy the MCP URL
5. Add to your IDE config:

{
  "mcpServers": {
    "zoho-mail": {
      "url": "https://mcp.zoho.com/api/v1/<your-server-id>/sse",
      "disabled": false,
      "autoApprove": ["send_mail"]
    }
  }
}

🚀 Usage

Quick Mode

# Compose standup email
daily-standup "Nikesh: Working on testing solution. No blockers. Aboli: Revamping SecurityBoat.net. No blockers. Sai: Researching Ollama models. No blockers." \
  --to [email protected] \
  --cc [email protected] \
  --sender "Mayur"

Interactive Mode

daily-standup interactive \
  --to [email protected] \
  --cc [email protected] \
  --sender "Mayur" \
  --time "10:40 – 11:50"

View System Prompt

daily-standup prompt

Options

| Flag | Description | Default | |------|-------------|---------| | --to <emails> | To recipients (comma-separated) | From .env | | --cc <emails> | CC recipients (comma-separated) | From .env | | --sender <name> | Sender name for signature | From .env | | --output <file> | Save email content to a file | — | | --model <model> | Groq model to use | llama-3.3-70b-versatile | | --time <time> | Meeting time (interactive mode) | — |

📧 Output

The agent outputs:

1. Human-readable email preview — the full MOM email in plain text
2. Zoho Mail MCP payload — structured JSON ready for the send_mail tool

The IDE agent router reads the MCP payload and calls:

• MCP Server: zoho-mail
• Tool: send_mail
• Arguments: the JSON payload from the agent output

📁 Project Structure

daily-standup-agent/
├── bin/
│   └── cli.js          # CLI entry point
├── src/
│   ├── index.js        # Main agent runner
│   ├── graph.js        # LangGraph pipeline
│   ├── state.js        # Agent state (Annotation)
│   ├── nodes.js        # Pipeline nodes
│   ├── prompts.js      # Email composition prompts
│   ├── llm.js          # Groq LLM configuration
│   └── setup.js        # Interactive setup
├── package.json
└── README.md

🤝 Programmatic Usage

const { runAgent } = require("daily-standup-agent");

const result = await runAgent(
  "Nikesh: Testing. Aboli: Design. Sai: Research.",
  {
    to: ["[email protected]"],
    cc: ["[email protected]"],
    sender: "Mayur",
    model: "llama-3.3-70b-versatile",
  }
);

console.log(result.output);      // Human-readable email
console.log(result.mcpPayload);  // Zoho Mail MCP payload JSON
console.log(result.sendStatus);  // "ready"

// The IDE agent router sends via:
// MCP Server: zoho-mail → Tool: send_mail → Args: result.mcpPayload

📄 License

MIT