← 返回 Skills 市场
1867
总下载
3
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install querit-search
功能描述
Web search via Querit.ai API. Use when you need to search the web for documentation, current events, facts, or any web content. Returns structured results with titles, URLs, and snippets.
安全使用建议
This skill appears to do what it says: it needs only your Querit API key and performs searches against Querit.ai and (optionally) fetches pages to extract readable content. Before installing: (1) Prefer inspecting files locally rather than piping a remote install script into bash — clone the repo and run npm ci yourself. (2) Be aware npm will install dependencies from the public registry; if you need stronger guarantees, audit the dependency tree or run it in a sandbox. (3) Avoid asking the skill to fetch internal or sensitive URLs (e.g., 169.254.x.x, 127.0.0.1, internal hostnames) because content.js will make HTTP requests from your environment and could disclose internal content to the agent. (4) Only provide a Querit API key you are comfortable using with this skill and consider creating a limited/revocable key if Querit supports that. If you want, I can point out the exact lines to edit if you prefer to remove the curl|bash installer or restrict content.js to a safe list of hosts.
功能分析
Type: OpenClaw Skill
Name: querit-search
Version: 0.1.0
The OpenClaw skill 'querit-search' is designed for web search and content extraction via the Querit.ai API. All analyzed files (SKILL.md, install.sh, content.js, search.js, package.json, _meta.json, README.md) show behavior consistent with this stated purpose. The `install.sh` script downloads skill files and installs npm dependencies to the standard OpenClaw directory. The `search.js` script makes authenticated requests to `api.querit.ai` using the `QUERIT_API_KEY` (as declared) and internally executes `content.js` via `node:child_process.execFile` to fetch and parse web pages. There is no evidence of intentional harmful behavior such as unauthorized data exfiltration, persistence mechanisms, or prompt injection attempts against the agent to perform actions beyond the skill's scope. All network and file system interactions are directly tied to the skill's functionality.
能力评估
Purpose & Capability
Name/description align with what the code does. The skill requires a single credential (QUERIT_API_KEY) which is used as a Bearer token to call https://api.querit.ai/v1/search. Search and content-extraction functionality is implemented in search.js and content.js respectively, which is appropriate for the described purpose.
Instruction Scope
SKILL.md and the CLI only instruct the agent to call the Querit API and (optionally) fetch and extract page content. That matches the stated purpose. A notable operational detail: content.js will fetch arbitrary URLs from the host where the skill runs and return extracted page content. This is expected for a content-extraction feature, but it means the skill can be used to load internal or protected endpoints (SSRF-like risk) if asked to fetch internal URLs. The instructions do not read unrelated env vars or config files.
Install Mechanism
The install flow is Node/npm-based (package.json + package-lock) and an install.sh that downloads files from raw.githubusercontent.com/interskh/querit-search/main and runs npm ci/install. GitHub raw and npm are common package hosts, but running a remote curl | bash installer and installing npm dependencies carries moderate risk (dependencies can include lifecycle scripts). The install script itself is straightforward and writes into ~/.openclaw/skills/querit-search; no obscure external downloads or archives are used beyond npm and GitHub raw.
Credentials
Only QUERIT_API_KEY is required and used as the primary credential for the Querit API. The skill documents alternative config/storage in OpenClaw config or .env; there are no unrelated or excessive environment variable requirements.
Persistence & Privilege
always is false. The skill installs under the user's skills directory and does not modify other skills or system-wide settings. It can be invoked autonomously by the agent (platform default), which is expected for a search skill; consider the normal autonomy considerations when granting runtime capability to call external APIs and fetch URLs.
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 querit-search - 安装完成后，直接呼叫该 Skill 的名称或使用
/querit-search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
v0.1.0
Querit Search skill initial release: Search the web and extract content using Querit.ai API.
- Search the web for documentation, facts, or current events and receive structured results with titles, URLs, and summaries.
- Flexible filtering: by language, country, date, included/excluded domains, and result count (up to 100).
- Optional extraction of full page content as markdown, or raw JSON output of results.
- Dedicated script for extracting readable content from a web page.
- Requires QUERIT_API_KEY environment variable; free tier available.
- Current limits: 72-character query max, 100 results per search, 1,000 queries/month.
元数据
常见问题
Querit Search 是什么？
Web search via Querit.ai API. Use when you need to search the web for documentation, current events, facts, or any web content. Returns structured results with titles, URLs, and snippets. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 1867 次。
如何安装 Querit Search？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install querit-search」即可一键安装，无需额外配置。
Querit Search 是免费的吗？
是的，Querit Search 完全免费（开源免费），可自由下载、安装和使用。
Querit Search 支持哪些平台？
Querit Search 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 Querit Search？
由 Kyle Sun（@interskh）开发并维护，当前版本 v0.1.0。
推荐 Skills