agentmail

What This Skill Does

AgentMail provides your OpenClaw AI agent with a dedicated, programmatic email inbox under the @agentmail.to domain. This skill empowers your agent to autonomously send, receive, and manage correspondence without human intervention. By integrating with the AgentMail API, the agent acts as an independent email client, capable of handling inbound inquiries, sending notifications, or participating in email-based workflows. It serves as a bridge between your agent's digital logic and the traditional, human-centric world of email.

Installation

Installation is a two-part process. First, request an API key from your human administrator via the AgentMail console at https://console.agentmail.to. Once obtained, install the Python SDK by running pip install agentmail. Configure your credentials securely by creating a JSON configuration file at ~/.agentmail/config.json with strictly restricted permissions (chmod 600). The OpenClaw platform can then interact with the local SDK to execute mail operations securely.

Use Cases

AgentMail is ideal for agents tasked with customer support, automated project management, or long-running tasks that require notification upon completion. An agent might monitor an inbox for client tickets, trigger automated responses based on data analysis, or send periodic status reports to stakeholders. It is also highly effective for "wake-up" triggers, where an agent starts a process upon receiving a specific email command.

Example Prompts

1. "Check my inbox for any emails from the accounting department and summarize the pending invoice requests for me."
2. "Send an email to [email protected] with the current status of the data scraping task and attach the summary report."
3. "Watch for any incoming emails with the subject 'Urgent' and notify me immediately if you find one."

Tips & Limitations

Always ensure your API key is stored with restricted file system permissions to prevent unauthorized access. Note that AgentMail is designed for agent-to-human or agent-to-agent communication; avoid sending high-volume bulk mail to prevent flagging by spam filters. Ensure your agent is configured to handle rate limits appropriately to maintain stability. Remember that the security of your inbox relies on the security of the configuration file; treat the config.json as sensitive credentials.