Skip to main content
Glama
cwattsnogueira

AVA

by cwattsnogueira
OverviewSchemaRelated ServersScoreDiscussions
Python
Local

AVA MCP Server – Gmail Draft Tool

This project implements a minimal Model Context Protocol (MCP) server named AVA, exposing a single tool: write_email_draft.

The tool creates Gmail drafts from structured input, allowing AI agents like Claude to generate email content safely — without sending anything. This enables reviewable automation for follow-ups, summaries, or outreach.

It integrates with the Gmail API using OAuth2 authentication, ensuring secure access and user control. The server responds with metadata including the draft ID, making it ideal for real-world AI workflows.

Architecture Overview

Claude AI → MCP Tool Call → AVA Server → Gmail API → Draft Created

┌────────────────────────────┐
│        Claude AI           │
│       (MCP client)         │
└────────────┬───────────────┘
             │  Tool Call: write_email_draft
             ▼
┌────────────────────────────┐
│      MCP Server (AVA)      │
│  mcp-server-draft.py       │
│  └── gmail.py              │
│  └── .env config           │
└────────────┬───────────────┘
             │  Uses credentials from .env
             ▼
┌────────────────────────────┐
│  Gmail API (OAuth2 Auth)   │
│  via Google Cloud Console  │
│  Project: ava-mcp-475222   │
└────────────┬───────────────┘
             │  Creates draft
             ▼
┌────────────────────────────┐
│     Gmail Draft Created    │
│                            │
└────────────────────────────┘

  • Claude AI: Sends structured tool calls via MCP

  • AVA Server: Receives calls, loads credentials from .env, and handles logic via gmail.py

  • Google Cloud Console: Hosts OAuth2 credentials (credentials.json)

  • Gmail API: Creates draft and returns metadata

Related MCP server: Gmail MCP Server

Features

  • MCP-compliant server using FastMCP

  • Gmail API integration via OAuth2

  • Secure credential handling via .env

  • Compatible with Claude and MCP Inspector

Project Structure

mcp-email-draft/
├── .env                    # Gmail credentials and sender email
├── mcp-server-draft.py     # MCP server exposing the tool
├── gmail.py               # Gmail API logic
├── pyproject.toml         # Project metadata and dependencies
├── credentials.json       # OAuth2 client credentials
├── token.json             # Gmail API token
├── README.md              # Documentation

Setup Instructions

1. Create project folder

cd C:/Users/carll/Desktop
mkdir mcp-email-draft
cd mcp-email-draft

2. Install uv

curl -Ls https://astral.sh/uv/install.sh | sh

3. Install dependencies

uv pip install -r pyproject.toml

Or manually:

uv pip install google-api-python-client google-auth-oauthlib python-dotenv mcp[cli]

Gmail API Setup

  1. Go to Google Cloud Console

  2. Create OAuth credentials for a desktop app

  3. Download credentials.json and place it in the project root

  4. Run the server once to trigger OAuth flow and generate token.json

.env File

USER_EMAIL=carlloswattsnogueira@gmail.com
GOOGLE_CREDENTIALS_PATH=C:/Users/carll/Desktop/mcp-email-draft/credentials.json
GOOGLE_TOKEN_PATH=C:/Users/carll/Desktop/mcp-email-draft/token.json

Claude Integration

Go to Claude Connectors and use:

{
  "mcpServers": {
    "AVA": {
      "command": "C:/Users/carll/.local/bin/uv.exe",
      "args": [
        "--directory",
        "C:/Users/carll/Desktop/mcp-email-draft",
        "run",
        "mcp-server-draft.py"
      ]
    }
  }
}

Usage Example

{
  "recipient_email": "carllos@example.com",
  "subject": "Follow-up on our meeting",
  "body": "Hi Carllos, thank you again for your time today. Looking forward to the next steps!"
}

Claude will return the Gmail draft ID and metadata. Next: Using Discord

License

This project is for demonstration and evaluation purposes only.
No affiliation with Google or Anthropic.

Available Tools

1 tool
write_email_draftC

Create a Gmail draft using recipient, subject, and body.

ParametersJSON Schema
NameRequiredDescriptionDefault
recipient_emailYes
subjectYes
bodyYes

TDQS

C2.4/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description must disclose behavioral traits. It only states the action without mentioning side effects (e.g., draft saved to Gmail, not sent), required permissions, or return behavior. The agent is left unaware of these critical factors.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single sentence, making it concise. However, it sacrifices necessary detail for brevity, earning a mid-range score for structure given the lack of informative content.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the absence of output schema and annotations, the description is incomplete. It fails to explain return values, error conditions, or confirmation of draft creation, leaving the agent without full context for invoking the tool.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters1/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 0%, and the description does not add meaning beyond the parameter names. It does not clarify expected formats for email, subject, or body, nor any constraints like character limits or encoding.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the action ('Create a Gmail draft') and the required inputs (recipient, subject, body). It distinguishes the tool's purpose as creating a draft rather than sending, though no sibling tools exist for comparison.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

No guidance is provided on when to use this tool versus alternatives, nor any conditions for use. The description lacks context about prerequisites, such as authentication or the difference between creating a draft and sending an email.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.

  1. 1 tool updatev0.1.0
    • First observedwrite_email_draft
Install Server
RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

TDQS

C2.8/5.0
Disambiguation5/5

With only one tool, there is no possibility of confusion or overlap. The purpose is uniquely defined.

Naming Consistency5/5

The single tool name 'write_email_draft' follows a clear verb_noun pattern, demonstrating internal consistency.

Tool Count3/5

A single tool is borderline; while it performs a concrete task, the server feels overly thin for anything beyond a very narrow use case.

Completeness2/5

The tool only creates a draft with no support for listing, updating, deleting, or sending, leaving significant gaps in email management.

Maintenance

ActivityInactive
ResponsivenessNo issues

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

Related MCP Servers

  • Gmail MCP Server

    JBR90
    A
    license
    Not graded
    quality
    D
    maintenance
    Enables AI assistants to manage Gmail by reading unread emails with automatic classification, creating AI-generated draft replies, and saving drafts directly to Gmail through the Gmail API.
    215
    MIT

  • Gmail MCP Server

    JaviEzpeleta
    A
    license
    Not graded
    quality
    D
    maintenance
    Enables AI assistants to interact with Gmail through OAuth2 authentication, allowing users to list, search, read emails, and create drafts with a safety-first design that prevents accidental sends by default.
    215
    MIT

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/cwattsnogueira/mcp-server-draft'

If you have feedback or need assistance with the MCP directory API, please join our Discord server