菜单
Run agent tasks in isolated, sandboxed environments to prevent malicious or untrusted code from affecting the host system.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Plan, reason, and verify across multiple levels using multi-step reasoning, tree-of-thought, and structured agent workflows.
Detect and repair security vulnerabilities or operational errors in the agent's own code and environment without manual intervention.
Generate and apply complete design systems from text descriptions or visual examples using AI and modern tooling.
Adapt user interfaces automatically based on device, viewport, user preferences, and contextual signals.
Build and query knowledge graphs automatically to capture relationships between entities, concepts, and project artifacts.
Store, retrieve, and evolve contextual and procedural knowledge using vector search, memory layers, and active forgetting.
| name | secure-execution-environment |
| description | Run agent tasks in isolated, sandboxed environments to prevent malicious or untrusted code from affecting the host system. |
Category: Agent Security Priority: High
This skill ensures that all tasks executed by DevinOS run inside isolated, secure environments. It prevents malicious, buggy, or untrusted code from affecting the host system, exfiltrating data, or escaping its intended boundaries.
To protect the host system, user data, and other projects from code that the agent writes, installs, or interacts with. It provides defense in depth for every command, test, and external dependency.
Use this skill when:
Scenario: The agent needs to execute a user-provided Python script that installs packages.
Agent Action:
Scenario: The agent executes a downloaded shell script with bash install.sh on the host.
Agent Action:
