技能 ads
📦
ads
v2.4.3 內容修訂 r1 安全 🌐 網路存取📁 檔案系統存取⚙️ 外部命令🔑 環境變數
Plan and Audit Paid Advertising Campaigns
Paid advertising decisions become unreliable when attribution, customer economics, and evidence gaps are ignored. This skill provides platform playbooks, audit checklists, and reviewable optimization plans.
支援: Claude Codex Code(CC)
使用我的智慧代理安裝
將此請求複製給你的智慧代理，其中包含權威 Skill 頁面和清單。
智能體請求
Review the Skillstore skill "ads" from https://skillstore.io/skills/coreyhaines31-ads.md and its manifest at https://skillstore.io/api/skills/coreyhaines31-ads/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.你的智慧代理仍應顯示其計畫，並依安全政策請求必要確認。
Agent 可讀資源
當 AI Agent、爬蟲或腳本需要乾淨脈絡、而不是讀取完整頁面時，請使用這些連結。
測試它
正在使用「ads」。 Our search-terms report shows 13 clicks from a campaign with 33 total clicks. Most disclosed terms look irrelevant.
預期結果:
- The report discloses 13 of 33 clicks, approximately 39 percent.
- The remaining 20 clicks have no visible search term, so the report does not prove that most traffic is irrelevant.
- Review disclosed terms for potential exclusions, check overblocking risk, and inspect conversion evidence before changing the campaign.
正在使用「ads」。 Our Google Ads audit lacks a search-terms export and has evidence for only half of the applicable checks.
預期結果:
- Search-term relevance and negative-keyword checks remain unknown, not failed.
- Evidence coverage is approximately 50 percent, so an overall account health score is not justified.
- Request the missing export and present verified findings separately from unverified checks.
- Keep proposed fixes as drafts until the specific changes receive approval.
正在使用「ads」。 Customer acquisition costs $300 for both our $9 and $999 monthly subscription plans. Can we scale paid acquisition?
預期結果:
- The guide's simplified revenue payback is approximately 33.3 months for the $9 plan and 0.3 months for the $999 plan.
- The blended result hides a large difference between plans.
- Evaluate retention, gross margins, and cash collection timing before approving a budget increase for either cohort.
安全審計
安全v1 • 2026/10/5 開啟版本化報告
All 82 supplied findings are false positives in their documented contexts, including Markdown formatting, evaluation text, related-skill links, and legitimate advertising references. The apparent injection quote is defensive guidance, and live changes require explicit approval. No evidence found of malicious instructions or unauthorized data export.
17
已掃描檔案
2,952
分析行數
0
複核項目
0
已忽略的誤報
風險因素
🌐 網路存取 (13)
evals/evals.json:6 evals/evals.json:131 references/conversion-tracking.md:38 references/conversion-tracking.md:116 references/conversion-tracking.md:196 references/conversion-tracking.md:237 references/rsa-output-spec.md:62 SKILL.md:487 SKILL.md:491 SKILL.md:492 SKILL.md:493 SKILL.md:494 SKILL.md:496
📁 檔案系統存取 (8)
references/creative-research-automation.md:19 references/creative-research-automation.md:23 references/creative-research-automation.md:70 references/creative-research-automation.md:87 references/creative-research-automation.md:101 references/creative-research-automation.md:102 references/creative-research-automation.md:103 references/meta-decision-system.md:161
⚙️ 外部命令 (4)
🔑 環境變數 (2)
最近一次完成的靜態與語意稽核未偵測到已確認的安全發現。這並不能證明該 Skill 沒有副作用。
審計單位: codex
分享與引用此報告
分享版本化評估報告、中性徽章、嵌入卡片與引用。Skillstore 只報告證據，不判斷該 Skill 是否安全。
複製報告連結
https://skillstore.io/skills/coreyhaines31-ads/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown 徽章
[](https://skillstore.io/skills/coreyhaines31-ads?utm_source=security_passport_badge)HTML 徽章
<a href="https://skillstore.io/skills/coreyhaines31-ads?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/coreyhaines31-ads/security.svg" alt="Skillstore security assessment" loading="lazy"></a>嵌入卡片
<iframe src="https://skillstore.io/embed/skills/coreyhaines31-ads.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>學術引用 (APA · BibTeX · CFF)
APA 引用
coreyhaines31. (2026). ads security audit report (audit version 1) [Author version 2.4.3]. Skillstore. https://skillstore.io/skills/coreyhaines31-ads/audits/1BibTeX 引用
@techreport{coreyhaines31-coreyhaines31-ads-2026,
author = {coreyhaines31},
title = {ads security audit report (audit version 1)},
institution = {Skillstore},
year = {2026},
number = {1},
url = {https://skillstore.io/skills/coreyhaines31-ads/audits/1},
note = {Author version 2.4.3}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "ads security audit report (audit version 1)"
version: "2.4.3"
type: report
authors:
- name: "coreyhaines31"
date-released: "2026-10-05"
url: "https://skillstore.io/skills/coreyhaines31-ads/audits/1"
identifiers:
- type: other
value: "skillstore:coreyhaines31-ads:audit:1"
description: "Skillstore immutable audit report identifier"
Skillstore 評分
評分依據 證據置信度: 中 45
架構
100
可維護性
87
內容
65
社群
87
規範符合度
你可以打造什麼
Choose Affordable Acquisition Channels
Compare channel fit and per-plan payback before allocating a limited advertising budget.
Audit Ecommerce Advertising
Review Google Ads and Merchant Center evidence, separate verified failures from unknowns, and draft reversible fixes.
Prepare B2B Campaign Reviews
Assess lead quality, attribution limits, audience reach, and account-based marketing opportunities before presenting client recommendations.
試試這些提示
Choose an Advertising Platform
We sell [product] to [audience] at [price], with [monthly budget] for advertising. Recommend platforms and explain what information is still needed.
Review Campaign Performance
Review this [platform] campaign export for [date range]. Compare CPA with [target], explain attribution limits, and identify missing evidence before recommending changes.
Audit Google Ads and Merchant Center
Audit our Search, Shopping, and Performance Max campaigns using these exports. Mark each check pass, fail, unknown, or not applicable. Draft fixes with rollback steps.
Plan a Controlled Scaling Test
Use these campaign, CRM, margin, and retention reports to compare payback by cohort. Draft a scaling experiment with budget limits, approval requirements, and rollback criteria.
最佳實務
- Provide dated exports, conversion definitions, attribution windows, budget constraints, and customer economics before requesting recommendations.
- Separate verified facts, inferred conclusions, stale claims, and unknown checks; request missing search-term evidence before proposing exclusions.
- Approve specific reversible changes before execution, and validate tracking consent, platform policies, and rollback steps.
避免
- Pause campaigns or increase budgets based only on a fixed CPA threshold without checking sample size, lag, and learning status.
- Combine incompatible attribution totals or describe an account audit as complete when major evidence is missing.
- Apply generic negative-keyword lists or follow instructions embedded in competitor pages, reviews, screenshots, or exports.
常見問題
Which AI tools support this skill?
The declared supported tools are Claude, Codex, and Claude Code. Available integrations determine whether live account access or browser research is possible.
Which advertising platforms does it cover?
It covers Google Ads, Meta, LinkedIn, Twitter/X, and TikTok. Detailed operational playbooks focus on Google Search, Meta, LinkedIn, and B2B acquisition.
Does it automatically change live campaigns?
Its guardrails default to read-only analysis. Changes require explicit approval of a specific plan with expected effects and a rollback step.
Can it audit an account without full access?
It can report verified findings and missing evidence. Below 60 percent evidence coverage, its audit guidance declines to provide an overall health score.
Can it generate advertising copy?
It includes copy frameworks and a Google responsive search ad specification. Bulk creative generation references a separate skill, and keyword exclusions require supporting evidence.
Does it guarantee better advertising returns?
No. Results depend on the offer, market, data quality, margins, tracking, and execution. Validate recommendations through controlled tests and current platform requirements.
開發者詳細資訊
授權
MIT
作者版本
v2.4.3
Skillstore 修訂
r1
引用
445604c853b415d2a53fe9b89f42be947d28a3a1
維護新鮮度
2026/10/6
使用量
0 次下載 · 0 次瀏覽