Openclaw Cfg Safe Edit

OpenClaw 配置文件修改前的验证规则。用于在修改 openclaw.json 或任何配置文件之前查询文档确认配置项可用性，避免因 OpenClaw 版本更新导致配置不兼容的问题。

Security Scan

The skill's purpose (validate OpenClaw config edits) matches its instructions, but the runtime instructions contain hard-coded local paths and call local tooling without declaring those requirements, which is inconsistent and risky.

Purpose & Capability

The SKILL.md describes validating OpenClaw configuration edits, which fits the skill name. However, the instructions assume the presence of a locally installed OpenClaw package and CLI (e.g., files under /home/root1/.npm-global/lib/node_modules/openclaw and the 'openclaw' gateway command) while the skill metadata declares no required binaries or environment variables. That mismatch (calling a local CLI but not declaring it) is incoherent.

Instruction Scope

Instructions tell the agent to run shell commands that list and grep a specific absolute path (/home/root1/.npm-global/...) and to invoke 'openclaw gateway config.schema.lookup'. These actions require file-system access and a specific user layout. Hard-coded paths tied to a particular user account are unexpected for a generic skill and could lead to incorrect behavior or unintended access to user files. The skill also instructs consulting online docs, which is reasonable, but the local-path reliance is problematic.

Install Mechanism

This is an instruction-only skill with no install spec and no code files, which is low-risk from an installation perspective. Nothing is downloaded or written to disk by the skill itself.

Credentials

The skill declares no required env vars or credentials, yet its instructions implicitly require access to a user's filesystem and an installed 'openclaw' CLI. The lack of declared required binaries (openclaw) and the use of an absolute home-directory path are disproportionate and ambiguous. A well-scoped skill should declare required binaries or use portable references (e.g., $HOME or a discovery step).

Persistence & Privilege

The skill is not set to always: true and does not request persistent/system-wide settings. It does not modify other skills or agent configuration according to the provided files.

What to consider before installing

This skill's goal (validate OpenClaw config changes) is reasonable, but the SKILL.md hard-codes a local path (/home/root1/...) and assumes an 'openclaw' CLI without declaring it. Before installing: 1) ask the skill author to remove hard-coded user paths and use $HOME or a discovery step; 2) require the skill metadata to declare the 'openclaw' binary (or document it); 3) confirm the agent will only access intended config files (not arbitrary home dirs); and 4) test in a safe environment. If you do not control /home/root1 or do not have the openclaw CLI installed, do not enable autonomous invocation until these issues are fixed.

配置验证规则

触发条件

当用户请求修改以下内容时自动触发：

• openclaw.json
• config/*.json
• 任何 OpenClaw 配置文件

验证流程

1. 查询本地文档

在修改配置前，先查询文档目录：

# 列出文档结构
ls /home/root1/.npm-global/lib/node_modules/openclaw/docs

# 搜索相关配置项
grep -r "<配置项名称>" /home/root1/.npm-global/lib/node_modules/openclaw/docs/ --include="*.md"

2. 查阅配置 schema

查询配置 schema 确认字段定义：

# 查看配置 schema 目录
ls /home/root1/.npm-global/lib/node_modules/openclaw/docs/cli/

# 或使用 gateway 工具查询 schema
openclaw gateway config.schema.lookup <dot-path>

3. 验证后再修改

确认配置项：

• 存在于当前版本的文档中
• 用法正确
• 类型匹配

如果文档中没有该配置项：

• 查看在线文档: https://docs.openclaw.ai
• 或检查 OpenClaw 更新日志

常见配置项参考

• gateway.* - 网关配置
• plugins.* - 插件配置
• channels.* - 通道配置
• models.* - 模型配置
• storage.* - 存储配置