技能 data-agent-platform-expert

📦

data-agent-platform-expert

Name: data-agent-platform-expert
Author: 图灵数盟

内容修订 r1 安全 📁 文件系统访问⚙️ 外部命令🌐 网络访问

Design Data Agent Platforms

Organizations struggle to convert specialist data knowledge into reliable agent systems. This skill provides structured methods for requirements, capabilities, workflows, controls, and evaluation.

支持: Claude Codex Code(CC)

🥉 77 青铜

使用我的智能体安装

将此请求复制给你的智能体，其中包含权威 Skill 页面和清单。

智能体请求

Review the Skillstore skill "data-agent-platform-expert" from https://skillstore.io/skills/dhturing-data-agent-platform-expert.md and its manifest at https://skillstore.io/api/skills/dhturing-data-agent-platform-expert/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

你的智能体仍应展示其计划，并按安全策略请求必要确认。

Agent 可读资源

当 AI Agent、爬虫或脚本需要干净上下文，而不是读取完整页面时，请使用这些链接。

Markdown 详情 GET /skills/dhturing-data-agent-platform-expert.md 已签名 manifest GET /api/skills/dhturing-data-agent-platform-expert/manifest 已签名 lockfile GET /api/skills/dhturing-data-agent-platform-expert/lockfile

测试它

正在使用“data-agent-platform-expert”。 Design requirements for a data inventory agent used by enterprise data teams.

预期结果:

Priority: inventory ingestion, classification, quality checks, and review queues.
Human checkpoint: approve classifications and exceptions before publication.
Controls: authorization, data isolation, audit logs, and versioned rules.
Metrics: review acceptance, completion time, exception rate, and cost per inventory.

正在使用“data-agent-platform-expert”。 Plan how to convert expert documents into an agent knowledge base.

预期结果:

Separate policy facts, decision rules, procedures, templates, and examples.
Attach source, owner, effective date, jurisdiction, and review status.
Test retrieval coverage and require evidence citations in generated drafts.

正在使用“data-agent-platform-expert”。 Define boundaries for an automated data asset assessment workflow.

预期结果:

The agent may prepare evidence summaries and draft assessments. Qualified reviewers must approve valuations, legal conclusions, and final customer deliverables.

安全审计

安全

v1 • 2026/9/27 打开版本化报告

All 12 static findings are false positives caused by documentation syntax, author links, readable Chinese text, or installation examples. No executable commands, hidden-file operations, encoded payloads, prompt injection, or other intent-level security risks were found.

已扫描文件

467

分析行数

复核项

已忽略误报

风险因素

📁 文件系统访问 (2)

profile.md:38 profile.md:38

⚙️ 外部命令 (3)

SKILL.md:20 SKILL.md:26 SKILL.md:33

🌐 网络访问 (2)

SKILL.md:6 SKILL.md:12

最新一次完成的静态与语义审计未检测到已确认的安全发现。这并不能证明该 Skill 没有副作用。

审计者: codex

分享与引用此报告

分享版本化评估报告、中性徽章、嵌入卡片和引用。Skillstore 只报告证据，不判断该 Skill 是否安全。

打开版本化报告

复制报告链接

https://skillstore.io/skills/dhturing-data-agent-platform-expert/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown 徽章

[![Skillstore security assessment](https://skillstore.io/badges/skills/dhturing-data-agent-platform-expert/security.svg)](https://skillstore.io/skills/dhturing-data-agent-platform-expert?utm_source=security_passport_badge)

HTML 徽章

<a href="https://skillstore.io/skills/dhturing-data-agent-platform-expert?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/dhturing-data-agent-platform-expert/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

嵌入卡片

<iframe src="https://skillstore.io/embed/skills/dhturing-data-agent-platform-expert.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

学术引用 (APA · BibTeX · CFF)

APA 引用

图灵数盟. (2026). data-agent-platform-expert security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/dhturing-data-agent-platform-expert/audits/1

BibTeX 引用

@techreport{dhturing-data-agent-platform-expert-2026,
  author = {图灵数盟},
  title = {data-agent-platform-expert security audit report (audit version 1)},
  institution = {Skillstore},
  year = {2026},
  number = {1},
  url = {https://skillstore.io/skills/dhturing-data-agent-platform-expert/audits/1},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "data-agent-platform-expert security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
  - name: "图灵数盟"
date-released: "2026-09-27"
url: "https://skillstore.io/skills/dhturing-data-agent-platform-expert/audits/1"
identifiers:
  - type: other
    value: "skillstore:dhturing-data-agent-platform-expert:audit:1"
    description: "Skillstore immutable audit report identifier"

Skillstore 评分

评分依据证据置信度: 中

架构

可维护性

内容

社区

规范符合性

你能构建什么

Plan a Data Agent Platform

Create prioritized requirements, modules, delivery phases, and governance controls for a new platform.

Structure Expert Knowledge

Convert specialist documents into searchable knowledge entries, decision rules, and reusable templates.

Review Agent Controls

Define capability boundaries, human checkpoints, security requirements, audit evidence, and operating metrics.

试试这些提示

Identify Platform Requirements

Review [business goal] and [user needs]. List platform requirements, priorities, assumptions, and required human decisions.

Structure a Knowledge Collection

Transform [document set] into knowledge entries, rules, templates, metadata, and an update process. Flag missing evidence.

Design an Agent Workflow

Design a workflow for [task]. Define inputs, reasoning stages, tool calls, outputs, controls, and human intervention points.

Create a Platform Delivery Plan

Create a phased plan for [platform]. Cover architecture, integrations, security, compliance, evidence, testing, operations, metrics, dependencies, and risks.

最佳实践

Provide business goals, target users, source documents, constraints, and required deliverables.
Verify every policy, cost, schedule, and compliance claim against current authoritative sources.
Keep human approval points for regulated decisions, exceptions, and final deliverables.

避免

Do not treat generated drafts as legal, valuation, or regulatory opinions.
Do not automate workflows before defining permissions, evidence, boundaries, and review ownership.
Do not estimate platform cost or duration without scope, integration, security, and staffing details.