技能 lifecycle-email
📦
lifecycle-email
内容修订 r1 中风险 🌐 网络访问⚙️ 外部命令📁 文件系统访问🔑 环境变量⚡ 包含脚本
Draft and Validate Opt-In Email Sequences
Generic follow-up emails can overlook the resource subscribers requested and the problem they described. This skill structures a three-part sequence and checks drafts locally.
支持: Claude Codex Code(CC)
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 "lifecycle-email" from https://skillstore.io/skills/cmj-hub-lifecycle-email.md and its manifest at https://skillstore.io/api/skills/cmj-hub-lifecycle-email/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。
测试它
正在使用“lifecycle-email”。 An opted-in trial user requested a checklist. Their problem is: the trial dies after the first login.
预期结果:
- Welcome: You asked for the trial checklist. Thank you for signing up.
- Bargain: Your requested trial checklist is available at [resource link].
- Nurture: You said the trial dies after the first login. Which step prevents users from returning?
正在使用“lifecycle-email”。 A draft contains a day 1 tip series and does not repeat the stated problem in nurture.
预期结果:
Validation refused the draft. Replace the generic series with problem-specific follow-up and include the complete stated pain phrase.
正在使用“lifecycle-email”。 A supplied draft contains incomplete JSON.
预期结果:
Input validation failed with an invalid JSON error. The scorer does not print the malformed draft.
安全审计
中风险v1 • 2026/10/3 打开版本化报告
Most static findings are false positives involving template literals, CSS, documentation, byte-order-mark handling, and legitimate local commands. The optional asset renderers contain HTML injection risks, external font requests, and destructive cleanup of a predictable directory. No evidence found of credential theft, malicious exfiltration, or prompt injection.
13
已扫描文件
711
分析行数
6
复核项
0
已忽略误报
已确认安全问题 (2)
中
assets/card.mjs:16-17assets/card.mjs:36-37assets/demo.mjs:11-15assets/demo.mjs:25-26assets/demo.mjs:67-68
Untrusted specification can terminate renderer HTML blocks
Both renderers interpolate accent_color directly into style blocks. The demo embeds JSON.stringify(s) directly into a script block without escaping HTML delimiters. A crafted specification can terminate either block and introduce executable browser content before any innerHTML assignment.
The source directly inserts unvalidated strings into HTML style and script contexts. JSON serialization does not escape HTML closing tags; exploitation requires rendering an untrusted specification.
中
Demo renderer deletes preexisting output-derived directories
The demo recursively deletes outGif plus the .frames suffix before creating its frame directory. It checks neither directory ownership nor prior contents. Selecting an output path can therefore erase an unrelated existing directory, and final cleanup repeats the recursive deletion.
rmSync uses recursive and force options on a predictable path derived from the caller's output argument. No exclusive temporary-directory creation or ownership check precedes deletion.
能力复核项 (6)
这些是真实的本地能力，对此技能可能是预期行为，因此需要复核，但不会按已确认恶意行为计分。
中
innerHTML assignment (XSS risk)
document.getElementById('lines').innerHTML = h;
Response lines from the supplied specification are concatenated into h without escaping. Assigning h to innerHTML allows attacker-controlled HTML and event handlers.
中
innerHTML assignment (XSS risk)
document.getElementById('artwrap').innerHTML = (ART && t>=T_ART)
Artifact labels and lines enter innerHTML without escaping. Rendering an untrusted specification can execute injected browser code.
低
Hardcoded URL
<link rel="preconnect" href="https://fonts.googleapis.com">
The generated page requests a preconnection to Google Fonts during rendering. This creates external network contact outside the offline scorer.
低
Hardcoded URL
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
The generated page preconnects to Google's font asset host. Rendering can expose connection metadata to an external service.
低
Hardcoded URL
<link href="https://fonts.googleapis.com/css2?family=Syne:wght@700&family=DM+Sans:opsz,wght@9..40,40
The stylesheet link loads Google Fonts while Chrome renders local assets. It exposes request metadata and prevents fully offline rendering.
低
Hardcoded URL
<link href="https://fonts.googleapis.com/css2?family=Syne:wght@700&family=DM+Sans:opsz,wght@9..40,40
The demo page loads a Google Fonts stylesheet in the browser. This exposes request metadata and introduces an external rendering dependency.
风险因素
🌐 网络访问 (6)
⚙️ 外部命令 (18)
assets/card.mjs:4 assets/card.mjs:97 assets/card.mjs:21-65 assets/card.mjs:65-66 assets/card.mjs:66-87 assets/card.mjs:95 assets/card.mjs:99 assets/card.mjs:101 assets/card.mjs:107 assets/demo.mjs:5 assets/demo.mjs:116 assets/demo.mjs:15-97 SKILL.md:31 SKILL.md:39-42 SKILL.md:42-46 SKILL.md:46 SKILL.md:52-59 tests/test_score.py:14
📁 文件系统访问 (7)
🔑 环境变量 (2)
⚡ 包含脚本 (2)
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。
复制报告链接
https://skillstore.io/skills/cmj-hub-lifecycle-email/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown 徽章
[](https://skillstore.io/skills/cmj-hub-lifecycle-email?utm_source=security_passport_badge)HTML 徽章
<a href="https://skillstore.io/skills/cmj-hub-lifecycle-email?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/cmj-hub-lifecycle-email/security.svg" alt="Skillstore security assessment" loading="lazy"></a>嵌入卡片
<iframe src="https://skillstore.io/embed/skills/cmj-hub-lifecycle-email.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
cmj-hub. (2026). lifecycle-email security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/cmj-hub-lifecycle-email/audits/1BibTeX 引用
@techreport{cmj-hub-cmj-hub-lifecycle-email-2026,
author = {cmj-hub},
title = {lifecycle-email security audit report (audit version 1)},
institution = {Skillstore},
year = {2026},
number = {1},
url = {https://skillstore.io/skills/cmj-hub-lifecycle-email/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: "lifecycle-email security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
- name: "cmj-hub"
date-released: "2026-10-03"
url: "https://skillstore.io/skills/cmj-hub-lifecycle-email/audits/1"
identifiers:
- type: other
value: "skillstore:cmj-hub-lifecycle-email:audit:1"
description: "Skillstore immutable audit report identifier"
Skillstore 评分
评分依据 证据置信度: 中 73
架构
85
可维护性
87
内容
65
社区
83
规范符合性
你能构建什么
Follow up after a resource request
Draft a welcome, deliver the requested checklist, and write follow-up copy addressing the subscriber's stated problem.
Support trial onboarding
Create an opt-in trial sequence focused on a specific onboarding problem, then check that all three sections are present.
Review client email drafts
Check client-provided drafts for missing sections, blocked phrases, and absent problem references before manual editorial review.
试试这些提示
Draft a first sequence
Draft an opt-in sequence for subscribers who requested [resource]. Their stated problem is [pain]. Include welcome, bargain, and nurture sections.
Improve an existing draft
Review [draft] for subscribers who opted in for [resource]. Separate welcome, bargain, and nurture. Keep the full pain phrase [pain] in nurture.
Validate a local draft
Check [draft_path] with scripts/score.py. Explain any refusal or input error and propose corrections. Do not send emails or run asset renderers.
Compare audience-specific sequences
Draft sequences for [audiences] using their documented consent, promised resources, and exact pain phrases. Validate each and compare messaging. Flag unsupported claims for review.
最佳实践
- Confirm subscriber consent and identify the exact resource promised before drafting.
- Use the complete stated problem phrase in nurture, then review the copy for clarity and accuracy.
- Run local validation on minimized drafts and manually review consent, links, claims, and delivery before sending.
避免
- Do not use this opt-in framework for unsolicited outreach.
- Do not replace problem-specific nurture with a generic calendar of tips.
- Do not treat a passing score as proof of consent, legal compliance, actual resource delivery, or campaign performance.
常见问题
Does this skill send email?
No. It guides drafting and validates text. Sending and scheduling require your email platform.
What does bargain mean here?
Bargain is the section that delivers the resource or access promised when the subscriber opted in.
What does the scorer require?
It requires four nonempty text fields: pain, welcome, bargain, and nurture. Nurture must contain the full pain phrase, ignoring case.
Why can a reasonable draft fail?
Configured phrase matches can reject otherwise reasonable copy. A paraphrased pain phrase also fails because matching requires the complete phrase.
Does validation need network access or paid APIs?
The Python scorer uses only the standard library and makes no network requests. Optional asset rendering loads external fonts.
Are the optional renderers safe for untrusted specifications?
No. They contain HTML injection risks, and the demo deletes a predictable frame directory. Avoid untrusted specifications and unnecessary renderer execution.