Engine (OpenClaw)
Tools & Skills
Built-in tools agents can invoke and skills that extend their capabilities.
Tools
Tools are typed functions that agents invoke during conversations. xCloud's engine comes with a comprehensive set of built-in tools.
Built-in Tools
| Tool | Function |
|---|---|
exec / process | Execute shell commands, manage processes |
code_execution | Sandboxed remote Python execution |
browser | Control Chromium (navigate, click, screenshot) |
web_search | Search the web |
web_fetch | Fetch web pages |
read / write / edit | File I/O in workspace |
apply_patch | Multi-hunk file patches |
message | Send cross-channel messages |
cron | Manage scheduled jobs |
image / image_generate | Analyze or generate images |
tts | Text-to-speech |
sessions_* | Session and sub-agent management |
Tool Groups
| Group | Includes |
|---|---|
group:runtime | exec, process, code_execution |
group:fs | read, write, edit, apply_patch |
group:web | web_search, x_search, web_fetch |
group:ui | browser, canvas |
group:automation | cron, gateway |
group:messaging | message |
group:media | image, image_generate, music_generate, video_generate, tts |
group:sessions | sessions_list, sessions_history, sessions_send, subagents |
Tool Profiles
Pre-configured sets of allowed tools:
| Profile | Description |
|---|---|
| full | No restrictions (default) |
| coding | File system, runtime, web, sessions, memory, cron, media |
| messaging | Only messaging and basic sessions |
| minimal | Only session_status |
Allow/Deny Lists
Fine-grained control over which tools agents can use:
{
"tools": {
"profile": "coding",
"allow": ["group:fs", "browser", "web_search"],
"deny": ["exec"]
}
}Deny always wins over allow.
Skills
Skills are Markdown files (SKILL.md) that get injected into an agent's system prompt. They provide context, constraints, and step-by-step guides.
Skill Location
Skills reside in the agent's workspace:
~/.openclaw/workspace/skills/
├── github/
│ └── SKILL.md
├── weather/
│ └── SKILL.md
└── code-review/
└── SKILL.mdRestricting Skills per Agent
{
"agents": {
"defaults": {
"skills": ["github", "weather"]
},
"list": [
{ "id": "writer" },
{ "id": "docs", "skills": ["docs-search"] },
{ "id": "locked-down", "skills": [] }
]
}
}Plugins
Plugins are packages that bundle channels, providers, tools, skills, and more. They run in-process with the gateway and should be treated as trusted code.
Core plugins ship with OpenClaw. External plugins are available via npm.