How do I install the Pwsh Execution skill?

Run npx skillmds@latest add xiongweilin/pwsh-execution in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Pwsh Execution safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Pwsh Execution?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Pwsh Execution free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Pwsh Execution?

xiongweilin (@xiongweilin) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Pwsh Execution by xiongweilin — Agent Skill

Name: Pwsh Execution
Author: xiongweilin

Pwsh Execution

Execute robust PowerShell workflows on Windows. Use when shell work needs quoting boundaries, multiline scripts, native exit-code handling, encoding discipline, or remote SSH calls from Windows. Do not use for pure file reads or read-only searches.

xiongweilin Updated Sep 17, 2026

File contents

PowerShell Execution

PowerShell is the agent's shell on Windows. Most failures here are quoting and state-boundary failures, not missing commands. This skill carries the executable method; the always-loaded invariants (check exit status before a consequential step, do not repeat an unchanged failed call) live in AGENTS.md.

Use when

Use for shell work that involves command construction, multi-line logic, native executables, encoding across Windows/Linux boundaries, or SSH from Windows.

Method

Quoting boundaries. Pass literal $... text through single-quoted strings. Delimit a variable before a literal colon as ${name}:. Keep PowerShell, SSH, and nested-shell quoting boundaries explicit; verify the resolved command before running it.

Multiline logic → script file. Prefer a here-string or a script file for cross-shell or multi-line scripts. Do not pipe CRLF-sensitive content into a remote shell. Use UTF-8 explicitly when text crosses Windows and Linux boundaries. Multi-line logic inside remote commands goes into a script file first; do not nest shell quoting inside inline commands.

Native exit codes. Capture the native command exit status and check it before a consequential next step. Treat expected no-match and nonzero native-command results explicitly ($LASTEXITCODE). A missing check turns a silent failure into a wrong next step.

Structured error handling. Treat expected failures as data: check for them, branch on them, and leave failure visible. Do not silently record failure as success.

Remote SSH from Windows. Invoke the fixed wrapper entry from the environment fact owner. Remote commands are interpreted by the remote login shell (Linux): logic containing spaces, redirection, or quoting must not be packed into an argument array — write a remote script and invoke the script path instead. Declare expected non-zero exit codes explicitly.

Success signal

A command was constructed with explicit quoting and encoding boundaries, its exit status was checked before the consequential next step, and any failure stayed visible.

---name: pwsh-executiondescription: Execute robust PowerShell workflows on Windows. Use when shell work needs quoting boundaries, multiline scripts, native exit-code handling, encoding discipline, or remote SSH calls from Windows. Do not use for pure file reads or read-only searches.---# PowerShell ExecutionPowerShell is the agent's shell on Windows. Most failures here are quoting and state-boundary failures, not missing commands. This skill carries the executable method; the always-loaded invariants (check exit status before a consequential step, do not repeat an unchanged failed call) live in AGENTS.md.## Use whenUse for shell work that involves command construction, multi-line logic, native executables, encoding across Windows/Linux boundaries, or SSH from Windows.## Method**Quoting boundaries.** Pass literal `$...` text through single-quoted strings. Delimit a variable before a literal colon as `${name}:`. Keep PowerShell, SSH, and nested-shell quoting boundaries explicit; verify the resolved command before running it.**Multiline logic → script file.** Prefer a here-string or a script file for cross-shell or multi-line scripts. Do not pipe CRLF-sensitive content into a remote shell. Use UTF-8 explicitly when text crosses Windows and Linux boundaries. Multi-line logic inside remote commands goes into a script file first; do not nest shell quoting inside inline commands.**Native exit codes.** Capture the native command exit status and check it before a consequential next step. Treat expected no-match and nonzero native-command results explicitly (`$LASTEXITCODE`). A missing check turns a silent failure into a wrong next step.**Structured error handling.** Treat expected failures as data: check for them, branch on them, and leave failure visible. Do not silently record failure as success.**Remote SSH from Windows.** Invoke the fixed wrapper entry from the environment fact owner. Remote commands are interpreted by the remote login shell (Linux): logic containing spaces, redirection, or quoting must not be packed into an argument array — write a remote script and invoke the script path instead. Declare expected non-zero exit codes explicitly.## Success signalA command was constructed with explicit quoting and encoding boundaries, its exit status was checked before the consequential next step, and any failure stayed visible.

Frequently asked questions