Initial commit: OpenClaw Skill Collection
6 custom skills (assign-task, dispatch-webhook, daily-briefing, task-capture, qmd-brain, tts-voice) with technical documentation. Compatible with Claude Code, OpenClaw, Codex CLI, and OpenCode.
This commit is contained in:
48
openclaw-knowhow-skill/docs/reference/cli/nodes.md
Normal file
48
openclaw-knowhow-skill/docs/reference/cli/nodes.md
Normal file
@@ -0,0 +1,48 @@
|
||||
# nodes
|
||||
|
||||
# `openclaw nodes`
|
||||
|
||||
Manage paired devices and enable invocation of node capabilities.
|
||||
|
||||
## Overview
|
||||
|
||||
The `openclaw nodes` command manages paired devices and enables invocation of node capabilities. This includes listing nodes, approving pending connections, checking status, and executing commands remotely.
|
||||
|
||||
## Key Commands
|
||||
|
||||
**Listing and Status:**
|
||||
|
||||
```bash
|
||||
openclaw nodes list
|
||||
openclaw nodes list --pending
|
||||
openclaw nodes list --connected
|
||||
openclaw nodes status <id|name|ip>
|
||||
```
|
||||
|
||||
The tool provides commands to display pending and paired nodes, with options to filter by connection status or timeframe. Users can retrieve only currently-connected nodes or those connecting within specified durations like 24 hours or 7 days.
|
||||
|
||||
**Execution:**
|
||||
|
||||
Remote command execution is available through two approaches:
|
||||
|
||||
```bash
|
||||
openclaw nodes invoke <id|name|ip> --method <method> --params <json>
|
||||
openclaw nodes run <id|name|ip> -- <command>
|
||||
```
|
||||
|
||||
The `invoke` command uses structured parameters, while the `run` command uses shell-style syntax. The system supports both direct commands and raw shell strings.
|
||||
|
||||
## Important Features
|
||||
|
||||
**Exec-style Defaults:**
|
||||
The `nodes run` command mirrors standard execution behavior by reading configuration from `tools.exec.*` settings and implementing approval workflows before invoking system commands.
|
||||
|
||||
**Customization Options:**
|
||||
Users can specify working directories, environment variables, command timeouts, and security levels. Additional flags allow requiring screen recording permissions and agent-scoped approvals.
|
||||
|
||||
**Node Identification:**
|
||||
Commands accept node references by ID, name, or IP address for flexible targeting.
|
||||
|
||||
## Notes
|
||||
|
||||
A compatible node must advertise `system.run` capabilities, such as a macOS companion application or headless node host.
|
||||
Reference in New Issue
Block a user