AI Agent Hub
Back to plugins
🧰

dsh-wsl

Web Tools Updated 2026.09.05

Run the following command in DeepSeek Harness:

dsh plugin install XINY11451/dsh-wsl

Paste the following prompt into your AI chat to install this plugin:

Run dsh plugin install XINY11451/dsh-wsl in your terminal to install the plugin; the source repository is at https://github.com/XINY11451/dsh-wsl.

About this plugin

When building AI agents on Windows that need Linux capabilities, developers have typically resorted to hand-written .sh scripts or PowerShell wrappers around wsl.exe — a brittle approach plagued by quoting edge cases, variable-expansion surprises, and path-format mismatches. dsh-wsl removes that middle layer entirely, exposing three clean model-facing tool interfaces so an agent can invoke Linux operations with simple parameters and no separate script maintenance.

The plugin registers three core tools: wsl executes any Linux command and returns stdout/stderr annotated with [exit code], [killed by signal], and [output truncated] markers; wsl-path converts between Windows drive paths and WSL Unix paths in either direction, auto-detected or explicitly forced; wsl-env summarises the active distributions, kernel version, CPU count, memory, and disk usage so an agent can reason about the environment before acting. Built-in safeguards include a destructive-command guard, automatic rewriting of Windows paths to /mnt/ prefixes, reliable environment-variable passthrough, and capped output with disk-spilling for large streams.

It is aimed at DSH developers working on Windows who want their agents to run Linux commands, inspect system resources, perform deployment or debugging tasks, and do all of it without maintaining a parallel script layer or shell wrapper.

Use Cases

  • Run Linux commands from a Windows agent without hand-written scripts
  • Convert Windows drive paths and WSL Unix paths bidirectionally
  • Inspect WSL distros, kernel, CPU, memory, and disk usage at a glance

Best For

  • Developers building DSH agents on Windows
  • Teams needing Linux tasks without a parallel script layer
  • Engineers who use WSL as their development environment