← 返回 Skills 市场
Openclaw Admin Skill
作者
rendrag-git
· GitHub ↗
· v1.0.3
· MIT-0
115
总下载
2
收藏
0
当前安装
4
版本数
在 OpenClaw 中安装
/install openclawadmin
功能描述
Use when diagnosing, configuring, fixing, tuning, or setting up anything in OpenClaw — gateway not responding, channel silent, model failover issues, auth er...
安全使用建议
This skill appears coherent and appropriate for on-host OpenClaw troubleshooting, but keep these cautions in mind before installing:
- It is an instruction-only admin helper that expects to run local commands and read files under ~/.openclaw. Only install it if you trust the agent runtime to run CLI commands with access to those paths.
- The SKILL.md forbids opening secrets and sending config externally, but that is a behavioral rule — it is not automatically enforced by your system. Do not hand over secrets; approve any request to view or paste secret values manually.
- The metadata did not declare the openclaw CLI as a required binary even though the instructions call it; verify openclaw is present in your environment before relying on the skill.
- The skill's source in registry metadata is unknown; SKILL.md points to a GitHub repo. If provenance matters, review the upstream repo or the README yourself before trusting it.
If you want extra safety, run the skill in a restricted environment/user account with only the necessary filesystem access, or review the SKILL.md and any referenced docs yourself first.
功能分析
Type: OpenClaw Skill
Name: openclawadmin
Version: 1.0.3
The skill bundle is a comprehensive administrative toolkit for managing OpenClaw installations, providing the agent with CLI references and configuration maps. It contains explicit safety instructions and guardrails (SKILL.md) designed to prevent the agent from reading secrets, exfiltrating data, or accidentally corrupting configuration files. The instructions are well-aligned with the stated purpose of system diagnosis and maintenance, emphasizing backups and validation without any evidence of malicious intent or harmful prompt injection.
能力标签
能力评估
Purpose & Capability
The name/description align with the instructions: troubleshooting and editing OpenClaw config, inspecting logs, and running the openclaw CLI. Minor mismatch: the SKILL.md assumes the openclaw CLI and local OpenClaw installation exist (refs to `<npm-prefix>/bin/openclaw`, local docs path), but the registry metadata did not declare a required binary. This is plausibly benign (it's an instruction-only skill) but worth noting so you ensure the runtime has the CLI available.
Instruction Scope
SKILL.md is explicit about rails: read-only diagnostics allowed, explicit user confirmation required for destructive actions, backup before edits, and an explicit prohibition on opening secrets files or transmitting secrets externally. It does permit fetching public docs (web fetch) for convenience; that is expected for an admin helper. The instructions do not instruct wide-ranging data collection or exfiltration.
Install Mechanism
No install spec and no code files are included (instruction-only). That minimizes disk-write / remote-install risk.
Credentials
The skill declares no required environment variables or credentials. It references local OpenClaw paths (e.g., ~/.openclaw/) which is appropriate for an admin tool. The SKILL.md explicitly forbids opening secrets files and sending them externally.
Persistence & Privilege
always is false and model invocation is allowed (default). The skill does not request persistent presence or modify other skills’ configuration. It asks for user confirmation before destructive or sensitive actions.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install openclawadmin - 安装完成后，直接呼叫该 Skill 的名称或使用
/openclawadmin触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.3
Gate reads of session/workspace file contents on explicit user approval (listing stays unrestricted). Session bodies never leave the conversation. Addresses ClawHub scan feedback on agent session PII exposure.
v1.0.2
Add Scope & Safety section (local-only; never open secrets.json/.env or transmit externally; explicit confirmation gates for destructive ops) and GitHub source link. Addresses ClawHub security scan feedback.
v1.0.1
Update doc discovery for recent npm builds: point to online docs (docs.openclaw.ai), GitHub source tree, or local clone — recent npm builds ship only a small reference/ subset. Add 'openclaw --version' as the first diagnostic step. Generalize <npm-prefix> paths.
v1.0.0
Initial release of openclaw-admin.
- Diagnose, configure, and repair OpenClaw installations.
- Direct filesystem access for inspecting and editing config, secrets, and logs.
- Step-by-step diagnostic workflow and safe configuration editing practices.
- Extensive documentation indexing and search guidance.
- Common troubleshooting tips for Discord, authentication, secrets management, the gateway, and agents.
- Emphasizes backup, validation, and safe editing to prevent config loss.
元数据
常见问题
Openclaw Admin Skill 是什么？
Use when diagnosing, configuring, fixing, tuning, or setting up anything in OpenClaw — gateway not responding, channel silent, model failover issues, auth er... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 115 次。
如何安装 Openclaw Admin Skill？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install openclawadmin」即可一键安装，无需额外配置。
Openclaw Admin Skill 是免费的吗？
是的，Openclaw Admin Skill 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Openclaw Admin Skill 支持哪些平台？
Openclaw Admin Skill 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Openclaw Admin Skill？
由 rendrag-git（@rendrag-git）开发并维护，当前版本 v1.0.3。
推荐 Skills