技能 job-application-agent

📦

job-application-agent

Name: job-application-agent
Author: vaibhavarora14

内容修订 r1 高风险 🌐 网络访问⚙️ 外部命令📁 文件系统访问🔑 环境变量⚡ 包含脚本

Manage Job Applications Safely

Job searches require repeated assessment, form completion, and accurate tracking. This skill organizes authorized applications with verified candidate data and submission safeguards.

支持: Claude Codex Code(CC)

⚠️ 38 差

安装前请停下并请求确认。

请审查安装计划，并在更改文件前获得用户明确同意。

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 "job-application-agent" from https://skillstore.io/skills/vaibhavarora14-job-application-agent.md and its manifest at https://skillstore.io/api/skills/vaibhavarora14-job-application-agent/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.

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/vaibhavarora14-job-application-agent.md 已签名 manifest GET /api/skills/vaibhavarora14-job-application-agent/manifest 已签名 lockfile GET /api/skills/vaibhavarora14-job-application-agent/lockfile

测试它

正在使用“job-application-agent”。 Assess a senior backend role with remote work and an unclear sponsorship policy.

预期结果:

Decision: Ask. The role and work mode match, but work authorization remains unclear. Submission is paused until sponsorship is confirmed.

正在使用“job-application-agent”。 Check whether a previously seen role can be submitted.

预期结果:

Submission blocked. The canonical job URL matches an existing ledger entry, so the application is treated as a hard duplicate.

正在使用“job-application-agent”。 Review recent application outcomes.

预期结果:

The review summarizes unique submissions, mature outcomes, interview quality, rejection reasons, source performance, and proposed targeting changes for candidate approval.

安全审计

高风险

v1 • 2026/8/25 打开版本化报告

The audit confirmed 12 static findings and dismissed 388 scanner matches that were tests, validation syntax, fixed resource paths, or defensive encoding. Material risks include default external telemetry and job sharing, mutable automatic updates, and plaintext profile data passed in a macOS process argument. Static review was capped at 400/738 representative findings; omitted static matches are unconfirmed, so automatic publishing stays disabled until manual review.

已扫描文件

5,820

分析行数

复核项

已忽略误报

已确认安全问题 (3)

高

SKILL.md:14

Automatic Execution of a Mutable Updater

The workflow runs an updater from a hidden home directory and describes automatic npm checks. Updated code can differ from the audited package.

SKILL.md directly instructs execution of the managed updater and states that a background updater checks npm every hour by default.

高

scripts/telemetry-client.mjs:81-85 scripts/job-application.mjs:1420-1426 scripts/telemetry-schema.mjs:41-60

Default-Enabled Job Activity Telemetry

The CLI sends installation, command, company, title, domain, fit, submission, and outcome metrics to an external relay without affirmative opt-in.

The configuration defaults to enabled, the first new-user command may send immediately, and the event schemas enumerate detailed job-search activity.

高

scripts/source-community-client.mjs:24-25 scripts/source-community-client.mjs:261-278 scripts/job-application.mjs:1337-1343

Default-Enabled Sharing of Application Metadata

Confirmed applications automatically share exact public job URLs, company names, roles, channels, and discovery sources with an external relay.

The client defaults sharing to enabled, and ledger addition immediately invokes community synchronization with normalized job metadata.

能力复核项 (12)

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。

高

SKILL.md:14

Hidden file in home directory

At the beginning of each workflow, run the managed updater once when `~/.agents/job-application-agen

SKILL.md:14 instructs the agent to execute a managed updater from a hidden home directory. That mutable executable can replace audited behavior outside this package revision.

高

scripts/secret-store.mjs:73-75

Process exec

exec('security', ['add-generic-password', '-a', account, '-s', service, '-U', '-w', raw], {

scripts/secret-store.mjs:73 passes the complete plaintext candidate profile to the macOS security command as an argument. Process arguments can expose sensitive profile data to local process inspection.

中

SKILL.md:14

Hidden file access

At the beginning of each workflow, run the managed updater once when `~/.agents/job-application-agen

SKILL.md:14 instructs the agent to execute a managed updater from a hidden home directory. That mutable executable can replace audited behavior outside this package revision.

低

scripts/source-community-client.mjs:194

Fetch API call

let response = await this.fetch(`${this.endpoint}/v1/install`, { method: 'POST', headers: { 'content

scripts/source-community-client.mjs:194 sends relay credentials, source metadata, or job metadata to an external service. The behavior is disclosed but enabled without affirmative opt-in.

低

scripts/source-community-client.mjs:202

Fetch API call

response = await this.fetch(`${this.endpoint}/v1/install`, { method: 'POST', headers: { 'content-typ

scripts/source-community-client.mjs:202 sends relay credentials, source metadata, or job metadata to an external service. The behavior is disclosed but enabled without affirmative opt-in.

低

scripts/source-community-client.mjs:256

Fetch API call

const response = await this.fetch(`${this.endpoint}/v1/sources`, { method: 'POST', headers: { 'conte

scripts/source-community-client.mjs:256 sends relay credentials, source metadata, or job metadata to an external service. The behavior is disclosed but enabled without affirmative opt-in.

低

scripts/source-community-client.mjs:306

Fetch API call

const response = await this.fetch(`${this.endpoint}/v1/jobs`, { method: 'POST', headers: { 'content-

scripts/source-community-client.mjs:306 sends relay credentials, source metadata, or job metadata to an external service. The behavior is disclosed but enabled without affirmative opt-in.

低

scripts/telemetry-client.mjs:110

Fetch API call

const response = await this.fetch(`${this.endpoint}/v1/install`, { method: 'POST', headers: { 'conte

scripts/telemetry-client.mjs:110 sends installation credentials or structured job activity to the configured relay. This is real outbound data transmission, even though payload schemas limit direct identity fields.

低

scripts/telemetry-client.mjs:133

Fetch API call

let response = await this.fetch(`${this.endpoint}/v1/events`, { method: 'POST', headers: { 'content-

scripts/telemetry-client.mjs:133 sends installation credentials or structured job activity to the configured relay. This is real outbound data transmission, even though payload schemas limit direct identity fields.

低

scripts/telemetry-client.mjs:137

Fetch API call

response = await this.fetch(`${this.endpoint}/v1/events`, { method: 'POST', headers: { 'content-type

scripts/telemetry-client.mjs:137 sends installation credentials or structured job activity to the configured relay. This is real outbound data transmission, even though payload schemas limit direct identity fields.

低

scripts/source-community-client.mjs:16

Hardcoded URL

export const DEFAULT_SOURCE_COMMUNITY_ENDPOINT = process.env.JOB_APPLICATION_AGENT_SOURCE_COMMUNITY_

scripts/source-community-client.mjs:16 defines the third-party relay used for default analytics or community sharing. It is a real external destination controlled outside the installed skill.

低

scripts/telemetry-client.mjs:9

Hardcoded URL

export const DEFAULT_TELEMETRY_ENDPOINT = process.env.JOB_APPLICATION_AGENT_TELEMETRY_URL ?? 'https:

scripts/telemetry-client.mjs:9 defines the third-party relay used for default analytics or community sharing. It is a real external destination controlled outside the installed skill.

风险因素

🌐 网络访问 (50)

⚙️ 外部命令 (50)

📁 文件系统访问 (50)

🔑 环境变量 (50)

⚡ 包含脚本 (3)

scripts/secret-store.mjs:62 scripts/secret-store.mjs:73 scripts/secret-store.mjs:105

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/vaibhavarora14-job-application-agent/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown 徽章

[![Skillstore security assessment](https://skillstore.io/badges/skills/vaibhavarora14-job-application-agent/security.svg)](https://skillstore.io/skills/vaibhavarora14-job-application-agent?utm_source=security_passport_badge)

HTML 徽章

<a href="https://skillstore.io/skills/vaibhavarora14-job-application-agent?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/vaibhavarora14-job-application-agent/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

嵌入卡片

<iframe src="https://skillstore.io/embed/skills/vaibhavarora14-job-application-agent.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

vaibhavarora14. (2026). job-application-agent security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/vaibhavarora14-job-application-agent/audits/1

BibTeX 引用

@techreport{vaibhavarora14-vaibhavarora14-job-application-agent-2026,
  author = {vaibhavarora14},
  title = {job-application-agent security audit report (audit version 1)},
  institution = {Skillstore},
  year = {2026},
  number = {1},
  url = {https://skillstore.io/skills/vaibhavarora14-job-application-agent/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: "job-application-agent security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
  - name: "vaibhavarora14"
date-released: "2026-08-25"
url: "https://skillstore.io/skills/vaibhavarora14-job-application-agent/audits/1"
identifiers:
  - type: other
    value: "skillstore:vaibhavarora14-job-application-agent: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

Organize a Personal Job Search

Create a verified profile, import a resume, and track each authorized application from discovery through outcome.

Assess Technical Roles

Compare postings with documented skills, eligibility, seniority, location, and compensation requirements before applying.

Manage Authorized Application Batches

Process an approved set of roles while enforcing duplicate checks, submission gates, and attention queues.

试试这些提示

Set Up My Profile

Create my job-search profile from this local resume: [path]. Ask only for required facts that the resume does not provide.

Assess One Role

Assess this job posting: [URL]. Check eligibility and must-have evidence, then explain whether I should review, skip, ask, or exclude it.

Prepare an Authorized Application

Prepare my application for [URL]. Use only verified profile facts, stop at protected questions, and request approval before submission.

Run and Review a Batch

Process these authorized job URLs: [URLs]. Enforce all eligibility and duplicate gates, queue blockers, record confirmed submissions, and summarize outcome metrics.

最佳实践

Use a verified resume and provide candidate facts directly when required information is missing.
Review eligibility, required evidence, compensation, and the final form before every submission.
Disable telemetry and community sharing before use when external data transmission is not acceptable.

避免

Do not invent experience, authorization, compensation, or answers that the candidate did not provide.
Do not bypass duplicate gates, MFA, CAPTCHA, legal attestations, or browser confirmation requirements.
Do not grant routine autonomy without a clear destination, scope, and submission authorization.