proomt

Search

Search posts, papers, and topics

All posts

Freek Van der Herten1 min readintermediate

sbx: Sandboxed Claude, complete with PHP and tools

Read on freek.dev

Summary

This article announces a practical guide on sandboxing Claude AI code execution using Docker's new sbx sandbox, featuring a custom PHP setup and additional tools. It also covers the rationale for making sandboxing the default for AI coding agents.

  • A guide will cover running Claude code in Docker's sbx sandbox.
  • It will include a custom PHP setup and extra tools for open source work.
  • The guide will explain why sandboxing AI coding agents is the sensible default.

Engineers deploying AI coding agents need to understand secure execution environments to prevent potential vulnerabilities from generated code.

2/10
llmsecurityaiphpdockersandboxing

Related reading

  2. Blog: How to Build a DevOps Agent

    This blog walks through the open‑source DevOps Agent Kit, which lets Claude Code or Cursor act as a DevOps assistant by connecting to CloudBees Unify, Jira, and Slack via MCP servers and to GitHub via the CLI. It provides a repeatable setup (Docker, Node, env file) and seven slash commands for pipeline overview, triage, security, release readiness, flag management, CI health scoring, and Jira tic…

    Codeshipcloudbees.com10 min

  4. Quoting Thariq Shihipar

    Claude Code version 2.1.277 now supports AGENTS.md as a fallback for CLAUDE.md for project instructions. This feature is built on an upcoming "Claude Code mods" system, which will allow users to create custom project instruction configurations.

    Simon Willisonsimonwillison.net1 minrelease

  5. Your AI coding agent evaluation is only as good as its sandbox

    Evaluating AI coding agents requires a robust sandbox to prevent agents from retrieving answers from the environment, which can invalidate tests of internal knowledge. A correct answer doesn't guarantee a valid measurement if the agent accessed information it shouldn't have, highlighting the need to define sandboxes by information boundaries rather than just tool restrictions. Always review agent…

    Microsoft for Developersmicrosoft.com5 min

  6. From Individual Experiments to a Shared AI Practice: How Avant Scaled AI-Assisted Engineering

    Avant replaced many personal AI‑assistant setups with a single, organization‑wide Claude Code configuration that embeds the company’s coding standards, project‑tracking links, and documentation. Treated as an internal product, the setup is versioned, owned by delivery engineers, and iteratively improved. The shared tool speeds routine tasks (scaffolding, tests, migrations, docs) and enforces cons…

    Moove-itqubika.com2 min