A powerful integration for running Claude Code as a non-interactive coding agent on remote machines through the OpenClaw Node architecture.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install claude-code-noninteractive-in-node
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install claude-code-noninteractive-in-node using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Claude Code Remote (OpenClaw Node)?
Claude Code Remote is a specialized capability within the Openclaw Skills ecosystem that allows developers to run Claude Code as a background, non-interactive agent on remote infrastructure. It bridges the gap between a central OpenClaw Gateway and remote Nodes, enabling seamless task delegation to different environments. This skill is particularly valuable for complex coding operations where the agent needs to reside on a separate machine from the gateway, or when utilizing high-speed local filesystems for remote projects is a priority for performance.
By wrapping commands in a login shell, this skill ensures that all necessary environment variables and project paths are correctly resolved on the remote host. It provides a robust framework for managing permissions and security policies, ensuring that AI-driven code modifications are handled safely across your distributed network using Openclaw Skills.
Claude Code Remote (OpenClaw Node) Use Cases
- Delegating intensive coding tasks from a local machine to a high-performance remote server.
- Executing multi-agent workflows where different agents operate on isolated Openclaw Skills nodes.
- Automating code refactoring and analysis in specialized environments like WSL or remote Linux instances.
- Running AI-assisted code migrations on fast local filesystems to avoid network latency issues.
How Claude Code Remote (OpenClaw Node) Works
- The OpenClaw Gateway initiates an execution request targeting a specific remote node using the host=node parameter.
- The command is processed through a bash login shell (bash -lc) to ensure all remote configuration and API keys are loaded.
- Claude Code is invoked in non-interactive mode using the --bare flag to prevent hanging on interactive prompts.
- The agent analyzes the project files on the remote node and communicates with the AI API to determine necessary code changes.
- Results, logs, and modified files are synchronized back to the gateway or processed according to the specified output format like stream-json.
Claude Code Remote (OpenClaw Node) Setup
To get started with this skill within the Openclaw Skills framework, follow these configuration steps:
- Ensure your API credentials are exported in the remote machine's
~/.bashrcfile before the interactive guard:
# Place this at the top of ~/.bashrc
export ANTHROPIC_API_KEY=sk-...
- Configure the remote node to handle non-interactive requests by setting up a worktree for fast filesystem access:
git worktree add -b remote-task /home/user/project-worktree main
- Execute the remote task through the gateway using the following JSON configuration:
{
"command": "bash -lc 'cd <project> && claude --bare -p \"<task>\"'",
"host": "node",
"node": "<your-node-id>"
}
Claude Code Remote (OpenClaw Node) Data Schema & Taxonomy
The skill manages data across remote boundaries with the following structure:
| Data Component | Description |
|---|---|
| Project Path | Remote-native paths (e.g., /home/user/project) for maximum I/O performance. |
| API Credentials | Managed via remote environment variables loaded through bash -lc. |
| Output Formats | Supports standard text or stream-json for integration with other Openclaw Skills. |
| Permissions | Hierarchical levels: Read (1), Analyze (2), Edit (3), and Full (4). |
Claude Code Remote (OpenClaw Node) Advanced Features
- Support for --dangerously-skip-permissions to enable fully autonomous automation in trusted remote environments.
- Multi-turn task management with configurable max-turns to prevent runaway AI processes on remote nodes.
- Advanced logging through --verbose and --include-partial-messages for real-time monitoring of remote agent progress.
- Integration with git worktrees to maintain a clean main branch while Openclaw Skills agents work on isolated features.
SKILL.md
Loading
Related Openclaw Skills
Generation Maker
Transform text prompts and raw media into professional 1080p video clips using an AI-powered cloud rendering pipeline.
Honest Critic
A specialized anti-sycophancy protocol that forces AI agents to provide honest critique and push back on flawed assumptions instead of offering empty praise.
Daily Report Generator
An automated tool for generating structured Excel reports for transmission fault assessment based on customizable cycles and performance metrics.
Claude Code Local
A specialized skill that enables Anthropic's Claude Code to operate as a local, non-interactive subordinate agent for automated development tasks.
Video Face
A powerful AI-driven skill for executing seamless face swaps in video clips through simple text prompts.
Local Web Search
A privacy-first, model-agnostic web search and content extraction skill that prioritizes local engines and anti-bot stealth fetching.
Featured*
