EmailAgent for Openclaw

An intelligent email assistant that uses LangChain and OpenAI to draft and send emails with a mandatory human approval step.

jawadsadiq01
v1.0.0
Feb 9, 2026
0
1.9k
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install langchain-email-agent

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install langchain-email-agent using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is EmailAgent?

The EmailAgent is a robust component designed for the Openclaw Skills ecosystem, enabling developers to integrate sophisticated email workflows into their AI agents. By leveraging the power of GPT-4o-mini and LangChain, this skill goes beyond simple templates by interpreting natural language instructions to craft tailored messages.

Its primary value lies in its balance of automation and oversight. While the AI handles the creative heavy lifting of drafting, the integrated human-in-the-loop middleware ensures that no email is sent without explicit verification, making it a reliable choice for professional communication environments.

EmailAgent Use Cases

  • Automating personalized outreach and follow-up sequences based on CRM data.
  • Drafting formal meeting requests with specific tone and context requirements.
  • Managing large-scale internal communications that require a final human touch.
  • Providing a secure AI-driven email drafting service for non-technical team members.

How EmailAgent Works

  1. The agent accepts a data transfer object containing the recipient's information and specific drafting instructions.
  2. It utilizes the LangChain framework to process these inputs through the configured OpenAI model.
  3. The system generates a draft and triggers the human-in-the-loop middleware to pause execution.
  4. A human reviewer interacts with the draft to approve, modify, or reject the message.
  5. Once approved, the EmailTool executes the final send operation, while read-only tools operate without interruption.

EmailAgent Setup

To get started with this skill in your environment, configure your model settings and initialize the agent as shown below:

export OPENAI_MODEL="gpt-4o-mini"

In your TypeScript application, you can then instantiate the agent:

import { EmailAgent } from './email.agent';
const agent = new EmailAgent();

EmailAgent Data Schema & Taxonomy

The skill utilizes a structured DTO to ensure consistency across Openclaw Skills implementations:

Attribute Type Requirement Description
email string Required The destination email address
name string Required The recipient's full name
subject string Optional The subject line for the email
body string Optional Initial content or context for the draft
instructions string Optional Directives for the AI regarding tone and style

EmailAgent Advanced Features

  • Human-in-the-Loop middleware providing Approve, Edit, and Reject capabilities for all outbound traffic.
  • Granular tool control allowing read operations to bypass approval while keeping write operations secure.
  • Support for custom instructions to enforce brand voice and formal communication standards.
  • Easy integration with other Openclaw Skills for complex, multi-step agentic workflows.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*