← Back to Skills Marketplace
🔌

Datayes Web Search

by Datayes · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ Security Clean
77
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install datayes-web-search
Description
通用 AI 语义搜索技能,通过 Datayes gptMaterials/v2 API 执行结构化语义搜索。用户询问最新资讯、研报、公告、会议纪要、行业动态、指标背景材料时使用。优先通过仓库内 Python 脚本发起请求,不要手写 HTTP 请求,不要在 skill 中硬编码 token。
README (SKILL.md)

Datayes Web Search

前提条件

1. 获取 Datayes Token

访问 https://r.datayes.com/auth/login 登录 Datayes 账号,并在 Datayes 控制台获取可撤销的 API token。

2. 配置 Token

macOS / Linux:

export DATAYES_TOKEN='your-token'

Windows PowerShell:

$env:DATAYES_TOKEN = "your-token"

Windows CMD:

set DATAYES_TOKEN=your-token

直接使用

  • 只用 python3 scripts/datayes_web_search.py 发请求。
  • 只用 Python 标准库;不要额外安装依赖。
  • 从环境变量 DATAYES_TOKEN 读取 token。脚本会自动带上 Authorization: Bearer \x3Ctoken>
  • 真实业务接口 URL 会先校验协议与主机名,只允许 Datayes 受信任域名,避免把 token 发到非 Datayes 地址。
  • 默认 rewrite=false,保留用户原始查询,减少意外改写。

最小流程

  1. 先查看接口规格。
python3 scripts/datayes_web_search.py spec
  1. 直接搜索并输出前 15 条结果。
python3 scripts/datayes_web_search.py search "新能源车 渗透率 2026" --query-scope research
  1. 需要更多结果时显式放大返回条数。
python3 scripts/datayes_web_search.py search "美国关税政策 2026" --query-scope news --top 30

queryScope 参数

搜索范围
(不传) 综合全部类型(默认)
research 券商研报
announcement 上市公司公告
news 财经资讯/新闻
meetingSummary 电话会/调研纪要
indicator 宏观/行业数据指标
researchTable 研报图表

选择规则

  • 用户问新闻、政策变化、市场动态时,优先 news
  • 用户问卖方观点、行业深度、公司覆盖时,优先 research
  • 用户问公告原文或正式披露时,优先 announcement
  • 用户问调研纪要、电话会、专家交流时,优先 meetingSummary
  • 用户问宏观或行业指标背景材料时,可用 indicator
  • 用户没限定范围时,可不传 queryScope 做综合搜索。

输出要求

  • 直接输出 API 返回的原始 JSON 数组,不做二次改写或字段重组。
  • 搜索词尽量精确,涉及时效性问题时优先附带年份或月份。
  • 若需要跨多种范围搜索,可分别调用多次,再由上层任务决定如何合并结果。
Usage Guidance
This skill appears coherent: it will run the included Python script and send your DATAYES_TOKEN bearer token to Datayes endpoints (gw.datayes.com). Before installing, make sure you trust Datayes and are willing to provide an API token; use a revocable token and follow least-privilege practices. Because the script prints the raw JSON returned by the API, be aware that returned data may include sensitive content—avoid pasting that output into untrusted places. If you need stricter controls, review or run the included scripts locally to confirm behavior and consider limiting network access or using short-lived tokens.
Capability Assessment
Purpose & Capability
Name/description say it's a Datayes semantic search skill; the only required binary is python3 and the only required env var is DATAYES_TOKEN. Those are appropriate and proportional for calling Datayes APIs.
Instruction Scope
SKILL.md instructs the agent to run the included Python script, to read the token from DATAYES_TOKEN, and not to hardcode tokens. The script only builds a POST to Datayes gptMaterials/v2 and prints the raw JSON response. The instructions do not ask the agent to read unrelated files or send data to other endpoints. (Note: the script outputs raw API JSON, which may contain sensitive material from Datayes.)
Install Mechanism
No install spec; skill is instruction-only with included Python scripts. No third-party packages are pulled and the script uses only the Python standard library—low installation risk.
Credentials
Only DATAYES_TOKEN is required. That credential is directly needed for Authorization to Datayes and is justified by the skill's purpose. No unrelated secrets or config paths are requested.
Persistence & Privilege
always is false and the skill does not modify other skills or system config. It runs only when invoked and does not request permanent elevated presence.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install datayes-web-search
  3. After installation, invoke the skill by name or use /datayes-web-search
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
- Initial release of datayes-web-search skill. - Enables structured semantic search for news, research reports, announcements, meeting summaries, and industry indicators using the Datayes gptMaterials/v2 API. - Uses Python scripts only (no manual HTTP requests or hardcoded tokens); authentication via environment variable DATAYES_TOKEN. - Safeguards to send API tokens only to trusted Datayes domains. - Returns raw API JSON results without modification. - Supports queryScope parameter to refine searches by content type.
Metadata
Slug datayes-web-search
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Datayes Web Search?

通用 AI 语义搜索技能,通过 Datayes gptMaterials/v2 API 执行结构化语义搜索。用户询问最新资讯、研报、公告、会议纪要、行业动态、指标背景材料时使用。优先通过仓库内 Python 脚本发起请求,不要手写 HTTP 请求,不要在 skill 中硬编码 token。 It is an AI Agent Skill for Claude Code / OpenClaw, with 77 downloads so far.

How do I install Datayes Web Search?

Run "/install datayes-web-search" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Datayes Web Search free?

Yes, Datayes Web Search is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Datayes Web Search support?

Datayes Web Search is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Datayes Web Search?

It is built and maintained by Datayes (@datayes); the current version is v1.0.0.

💬 Comments