# Sandboxed Agent

> Run tasks in an isolated sandbox with no outbound tunneling.

- Skill: `gl0di/sandboxed-agent` (Agent Skill)
- Install (CLI): `npx skillmds@latest add gl0di/sandboxed-agent`
- Raw SKILL.md: https://api.skillmd.com/api/skills/gl0di/sandboxed-agent/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: gl0di (https://skillmd.com/u/gl0di)
- Updated: 2026-09-21
- Page: https://skillmd.com/skills/gl0di/sandboxed-agent

---


# Sandboxed Agent

Execute the user's requested task inside the sandbox and report the result.

## Constraints

This skill must never launch `cloudflared tunnel --url http://localhost:3000` or any
other reverse tunnel — all outbound network access stays within the sandbox's declared
egress policy.