Execute Python, Node.js, Go, and Bash scripts in a Docker-sandboxed environment.
npx skills add https://github.com/nielsmadan/agentic-coding --skill workbench-nielsmadan
Please help me install this Agent Skill. Skill: workbench Source: https://github.com/nielsmadan/agentic-coding/tree/main/claude/skills/workbench Command: npx skills add https://github.com/nielsmadan/agentic-coding --skill workbench-nielsmadan
This Skill provides a secure, isolated Docker environment for running code, preventing untrusted or experimental scripts from affecting your host system.
Use the workbench skill to build a python script that reads a CSV and outputs summary stats.
High-intent search queries and answers about installing and using this skill.
To run untrusted Python scripts in an isolated Docker environment, this skill executes your code within a sandboxed container, ensuring complete isolation from your host system. The environment persists across sessions, retaining installed packages for iterative development.
Yes, you can execute Node.js, Go, and Bash code inside a secure sandbox. This skill supports multiple languages for script development and execution within a Docker-sandboxed environment, preventing experimental scripts from affecting your host machine.
Yes, the Docker sandbox does persist installed npm packages across code execution sessions. The container manages its lifecycle to maintain a persistent environment, allowing iterative development and testing without losing previously installed dependencies or script states.
To safely test scripts that interact with external APIs without risking your local setup, execute them within this isolated Docker sandbox. It completely separates experimental code and package installations from your host machine, requiring user approval for initial build and run commands.
Yes, you need to manually approve Docker build commands when running code in a sandbox. The skill manages the container lifecycle, including setup, execution, and cleanup, but explicitly requires user approval for initial build and run commands to maintain security.
kurtosis-techBuilds and tests Kurtosis from source on local Docker, compiling Engine, Core, Files Artifacts Expander, CLI, tagging images with git short SHA, and restarting the engine for a clean test environment.
onebrain-aiPromote recurring insights from session logs into persistent memory files.
abnejLLCManage organizational policy lifecycles from drafting through approval and updates.
sebastiankunikowskiExecute untrusted code in isolated Deno sandboxes with Firecracker microVMs.
