← Back to Skills Marketplace
myskil
by
Jizhidemu52
· GitHub ↗
· v1.0.1
· MIT-0
82
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install webmail-email-monitor-skill
Description
Monitor Outlook and other common webmail inboxes in a persistent Edge profile, process new messages as a detached local background task, capture complete mes...
Usage Guidance
Do not run or allow unsigned PowerShell scripts you didn't review. Before installing or running this skill: (1) ask the publisher for the actual scripts and their cryptographic provenance (source repo, release URL, or a package you can inspect); (2) request an explicit install spec and a list of network hosts the bootstrap will contact; (3) review the contents of any scripts (especially those run with ExecutionPolicy Bypass) to ensure they don't download arbitrary code or persist startup tasks; (4) consider running in a sandboxed or VM environment first; (5) be aware this will access your browser profile, download email attachments to your Desktop, and keep a long-running background process — only proceed if you trust the code and understand those implications.
Capability Analysis
Type: OpenClaw Skill
Name: webmail-email-monitor-skill
Version: 1.0.1
The skill bundle (specifically SKILL.md) describes a background webmail monitor that employs high-risk persistence and execution tactics, including bypassing PowerShell execution policies and installing system-level software via winget. It instructs the agent to launch a detached background process and 'return immediately,' effectively hiding the monitor's ongoing activities—such as capturing screenshots and downloading attachments—from the user's immediate oversight. While these capabilities are framed as automation features, the combination of environment modification, sensitive data access, and stealthy execution patterns is highly suspicious.
Capability Assessment
Purpose & Capability
The skill claims to run a persistent webmail monitor via scripts under scripts/webmail_monitor_cli.ps1 and Python shims, but the published bundle contains only SKILL.md (no scripts, no code, no install). The SKILL.md expects local scripts, dependency installation, and a 'webmail-monitor' shim to exist or be bootstrapped, yet the package provides none and declares no required env/config. That incoherence means the skill as published cannot perform its stated purpose without fetching or creating code at runtime.
Instruction Scope
The instructions explicitly tell the agent to run PowerShell with ExecutionPolicy Bypass to start a detached background monitor, capture full message screenshots, download attachments to the desktop, and write logs and Excel files. Those actions require broad local file and browser profile access and have persistence implications. The SKILL.md gives concrete runtime commands (powershell -ExecutionPolicy Bypass -File scripts/...) even though those files are missing, which is a red flag for hidden fetch/exec behavior if the agent or user follows the instructions.
Install Mechanism
There is no install spec in the registry entry, but the SKILL.md says the CLI 'auto-runs a bootstrap step that installs missing Python dependencies and refreshes the local webmail-monitor shim.' That implies dynamic installation or downloading code at runtime, yet no trusted source/URL or release host is specified. The lack of included code plus an implied bootstrapping installer is a risky mismatch.
Credentials
The skill declares no required environment variables or config paths, but the instructions require access to an Edge profile, the user's Desktop, and persistent runtime files (runtime/monitor.log, desktop workbook/folders). Asking for none of these in metadata while instructing operations on sensitive local artifacts is inconsistent and prevents pre-install review of what credentials or profiles will be used.
Persistence & Privilege
The skill is explicitly designed to run as a detached, persistent background process on the user's machine and to keep an Edge profile open for mailbox login. Although the registry flags do not force 'always: true', the technical behavior described (ExecutionPolicy Bypass, background daemon, PID tracking, persistent profile reuse) gives it substantial local persistence. Combined with the missing scripts/install details, this raises risk of long-lived code running without clear provenance.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install webmail-email-monitor-skill - After installation, invoke the skill by name or use
/webmail-email-monitor-skill - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
- Improved documentation in SKILL.md with clear instructions, rules, and operating procedures for running the email monitor on Windows.
- Expanded explanations on default behaviors, Python-first boundaries, and the division of logic between scripts and prompts.
- Explicit guidance for handling mailbox login, workflow automation, and persistent Edge profiles.
- Listed required local output artifacts and enforced specific Excel row structure for processed emails.
- Added detailed prerequisites and reuse requirements for cross-machine compatibility.
- Provided standard command-line usage examples for starting, stopping, checking status, and bootstrapping the monitor.
Metadata
Frequently Asked Questions
What is myskil?
Monitor Outlook and other common webmail inboxes in a persistent Edge profile, process new messages as a detached local background task, capture complete mes... It is an AI Agent Skill for Claude Code / OpenClaw, with 82 downloads so far.
How do I install myskil?
Run "/install webmail-email-monitor-skill" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is myskil free?
Yes, myskil is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does myskil support?
myskil is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created myskil?
It is built and maintained by Jizhidemu52 (@jizhidemu52); the current version is v1.0.1.
More Skills