Manage cloud sandbox environments and sessions via the Runtm Cloud CLI.
npx skills add https://github.com/runtm-ai/runtm --skill runtm
Please help me install this Agent Skill. Skill: runtm Source: https://github.com/runtm-ai/runtm/tree/main/packages/agent/skills Command: npx skills add https://github.com/runtm-ai/runtm --skill runtm
AI agents require a programmatic interface to manage isolated sandbox environments, execute commands, handle files, and deploy applications without manual dashboard interaction. This Skill provides the complete Runtm Cloud CLI reference, enabling agents to perform any cloud operation autonomously.
Use the runtm skill to create a template from a GitHub repository, wait for it to become ready, launch a session from that template, and execute a verification command inside the sandbox.
High-intent search queries and answers about installing and using this skill.
Cloud sandbox provisioning for AI coding agents is automated via a comprehensive CLI that manages isolated environments, executes commands, and handles files without manual dashboard interaction. It orchestrates session lifecycles and creates templates from Git repositories.
Creating cloud sandbox templates from Git repositories involves cloning the repository into a template, pre-packaging dependencies and services, then building and snapshotting it. This allows coding agent sessions to inherit the correct context and credentials upon launch.
Yes, managing isolated cloud sandbox environments requires the runtm-api binary and valid API credentials to execute authenticated requests against the cloud control plane for provisioning, deploying, and manipulating file systems.
Yes, you can manage the complete lifecycle of coding agent sessions programmatically. The CLI supports creating, connecting, executing, prompting, pausing, resuming, and destroying sessions, enabling autonomous cloud operations and end-to-end workflows.
You attach MCP servers, skills, and tool providers to sandbox templates through integration management features. Defining and attaching these integrations ensures that launched coding agent sessions automatically inherit the correct external context, tools, and credentials.
The best way to debug AI coding agent failures in an isolated cloud sandbox is to launch a session, execute verification commands, and monitor live events. This allows you to identify failures, apply fixes, and open a PR with changes entirely via structured CLI commands.
aronprinsManage AI-agent companies via the Paperclip CLI from the terminal.
kartikkabadiExecute commands and code in isolated sandbox environments with preview URLs.
openstatusHQBuild MCP servers in Python or TypeScript with validated tool schemas.
Moisesjr20Run agent-browser inside ephemeral Vercel Sandbox microVMs for isolated browser automation.
