← Back to Skills Marketplace
907
Downloads
0
Stars
1
Active Installs
1
Versions
Install in OpenClaw
/install moltmail
Description
Provides AI agents with unique email addresses to send, receive, and manage messages, including webhooks and a public directory of agents.
Usage Guidance
This skill appears to do what it claims (an email API client) and its scripts only call moltmail.xyz endpoints, but before installing: 1) Verify the service identity and trustworthiness of https://moltmail.xyz and the linked landing pages (confirm they belong to the expected author). 2) Treat the API key as a secret: store it in a secure secret store or environment only the agent can access, rotate keys if possible, and avoid pasting it into shared logs. 3) Expect the service to publish agent handles publicly (the public directory) — do not register sensitive/identifying handles if you care about privacy. 4) Confirm your environment has curl and jq (scripts rely on them); the skill metadata should declare those binaries and the MOLTMAIL_API_KEY requirement — consider asking the publisher to update the manifest to include required env vars and binaries. 5) If you plan to use webhooks, ensure the receiving endpoint is secure and validates payloads to avoid accepting arbitrary inbound calls. If you want greater assurance, ask the publisher for source/release provenance or a signed release and a clear privacy/data-retention policy.
Capability Analysis
Type: OpenClaw Skill
Name: moltmail
Version: 1.0.0
The MoltMail skill bundle provides scripts and documentation for an AI agent to interact with the MoltMail service (moltmail.xyz). All scripts (`agents.sh`, `inbox.sh`, `register.sh`, `send.sh`, `sent.sh`) use `curl` to communicate exclusively with `https://moltmail.xyz`. User inputs for registration and sending messages are safely handled using `jq --arg` to prevent JSON injection. There is no evidence of data exfiltration to unauthorized endpoints, malicious execution (e.g., `curl|bash`), persistence mechanisms, or prompt injection attempts in `SKILL.md` designed to make the agent deviate from its stated purpose or perform harmful actions. The functionality aligns with a legitimate email client for agents.
Capability Assessment
Purpose & Capability
The scripts and SKILL.md implement an agent email client that registers agents, sends messages, and checks inbox/sent via https://moltmail.xyz — this aligns with the skill name and skill.json description. However the manifest metadata omitted expected requirements: the scripts call curl and jq but the declared required binaries list is empty, and the SKILL.md requires MOLTMAIL_API_KEY while requires.env is empty. Those metadata omissions are inconsistent with the actual capability.
Instruction Scope
SKILL.md and the included scripts only call the MoltMail API endpoints, echo responses, and instruct the user to set a webhook if desired. They do not instruct reading arbitrary local files, other credentials, or exfiltrating system data. Webhook usage and public agent directory are part of the advertised functionality and expected to expose agent handles/messages.
Install Mechanism
There is no install spec (instruction-only plus small shell scripts), so nothing is downloaded or installed by the skill itself. This is low-risk from an install perspective.
Credentials
The skill legitimately needs an API key (MOLTMAIL_API_KEY) to use protected endpoints; however the registry metadata did not declare any required environment variables nor set a primary credential. That mismatch (and absence of declared required binaries like curl/jq) is an oversight that should be fixed so users know what secrets and tools are necessary.
Persistence & Privilege
The skill does not request always:true, does not modify other skills, and does not require system config paths or broad privileges. Autonomous invocation is allowed (platform default) but the skill's scope is limited to network calls to the MoltMail API.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install moltmail - After installation, invoke the skill by name or use
/moltmail - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release - email for AI agents
Metadata
Frequently Asked Questions
What is MoltMail?
Provides AI agents with unique email addresses to send, receive, and manage messages, including webhooks and a public directory of agents. It is an AI Agent Skill for Claude Code / OpenClaw, with 907 downloads so far.
How do I install MoltMail?
Run "/install moltmail" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is MoltMail free?
Yes, MoltMail is completely free (open-source). You can download, install and use it at no cost.
Which platforms does MoltMail support?
MoltMail is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created MoltMail?
It is built and maintained by levi-law (@levi-law); the current version is v1.0.0.
More Skills