← 返回 Skills 市场
SQLBot
作者
xuwei-fit2cloud
· GitHub ↗
· v1.1.0
· MIT-0
112
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install sqlbot
功能描述
Manage SQLBot workspaces, datasources, ask-data flows, and dashboards, including listing and switching workspace or datasource context, asking questions agai...
安全使用建议
This skill appears to do what it says: it runs a local Python script that calls your SQLBot instance and saves a small local state file. Things to consider before installing:
- You must supply SQLBOT_BASE_URL and an API key pair in a .env next to the skill (or via --env-file). Only provide these credentials if you trust the SQLBot endpoint and are okay with the skill making signed requests on your behalf.
- The skill will write a state file (default .sqlbot-skill-state.json) into the skill directory (or another configured path). If you need isolation, change SQLBOT_STATE_FILE to a safe location.
- Exporting dashboards uses Playwright (a browser) and a derived frontend preview URL — the browser will send the same API auth headers to the SQLBot frontend. Install Playwright only if you trust this workflow.
- The registry metadata omits the required env vars. Confirm you’re comfortable creating a .env with secrets before enabling the skill. If you want extra assurance, review the full sqlbot_skills.py source locally or run it in an isolated environment/container.
能力评估
Purpose & Capability
The name/description match the included script and instructions: the skill is a CLI wrapper that lists/switches workspaces and datasources, asks questions, lists and exports dashboards. The script expects SQLBot base URL and API key pair in a .env file — that is appropriate for a service-integrating skill. One minor inconsistency: the registry metadata shows no required env vars, but the SKILL.md and the script clearly require SQLBOT_BASE_URL, SQLBOT_API_KEY_ACCESS_KEY and SQLBOT_API_KEY_SECRET_KEY; this is an omission in the registry metadata rather than a functional mismatch.
Instruction Scope
The SKILL.md tells the agent to run the bundled python script under ${CLAUDE_SKILL_DIR} and to check for and use a .env next to the skill. The script will read that .env and may be given a different --env-file. It will make HTTP calls to the configured SQLBot base URL and may invoke Playwright (browser) for exports. The skill also writes a small local state file (default .sqlbot-skill-state.json) to persist current workspace/datasource. It does not instruct the agent to read arbitrary unrelated system files or transmit data to unknown external endpoints — network activity is scoped to SQLBOT_BASE_URL (derived frontend preview URLs are deduced from that base URL).
Install Mechanism
There is no install spec; this is essentially an instruction-only skill that runs the included Python script. No external downloads or remote installers are referenced in SKILL.md. The only optional runtime dependency is Playwright for exports, which the SKILL.md instructs how to install via pip/playwright (a known tool).
Credentials
The script legitimately needs SQLBot credentials (base URL, access key, secret key) to call the service; those are requested via a .env file. That is proportionate to the stated functionality. As noted above, the skill registry entry did not declare those env vars as required, so the registry metadata underreports the secrets the skill needs. The script keeps credentials in process memory to sign API requests but does not appear to persist the secrets to disk (it does persist only the workspace/datasource state).
Persistence & Privilege
The skill writes a local state file (by default in the skill directory) to persist current workspace and datasource selection — this is reasonable for its workflow. It does not request always: true and does not modify other skills or global agent configs. SKILL.md sets disable-model-invocation: true (it will not invoke the model autonomously), which reduces risk.
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 sqlbot - 安装完成后，直接呼叫该 Skill 的名称或使用
/sqlbot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
v1.1.0
- Added support for exporting dashboards as JPG in addition to PNG and PDF formats.
- Updated documentation and argument hints to reflect new JPG export option and format handling.
- Clarified dashboard export default output path to use JPG unless another format is specified.
v1.0.0
Initial release of SQLBot Workspace Dashboard Skill.
- Provides commands to manage SQLBot workspaces, datasources, ask-data flows, and dashboards.
- Supports listing and switching workspace or datasource context.
- Allows asking questions against a datasource, listing dashboards, viewing details, and exporting dashboards as PNG or PDF.
- Includes prerequisite setup instructions and guidance for Playwright installation for exports.
- Offers clear mapping from user requests to command-line usage and best practices for context management.
元数据
常见问题
SQLBot 是什么？
Manage SQLBot workspaces, datasources, ask-data flows, and dashboards, including listing and switching workspace or datasource context, asking questions agai... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 112 次。
如何安装 SQLBot？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install sqlbot」即可一键安装，无需额外配置。
SQLBot 是免费的吗？
是的，SQLBot 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
SQLBot 支持哪些平台？
SQLBot 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 SQLBot？
由 xuwei-fit2cloud（@xuwei-fit2cloud）开发并维护，当前版本 v1.1.0。
推荐 Skills