Security audit

求职申请专业版

Security checks for vulnerabilities and agentic risk

Overview

This skill is a bulk job-application automation guide that can submit real applications at scale and includes under-scoped tracking, API, scheduling, and account-rotation guidance.

Review before installing. Only use this with explicit candidate consent, verified platform permission, dry-run review, strict per-application approval, secure credential handling, authenticated local-only API exposure, and disabled account-rotation or anti-detection behavior. Also inspect any referenced Python scripts before running them, because this package does not include the implementation it tells the agent to execute.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (7)

Vague Triggers

Medium

Confidence: 92% confidence
Finding: The activation/description text is unusually broad, covering generic efficiency, automation, workflow optimization, search, retrieval, and enterprise-team use cases rather than a narrowly scoped job-application function. In an agent setting, this can cause the skill to activate for unrelated requests and then gain access to file, search, and exec capabilities, increasing the chance of unintended external actions or data handling.

Missing User Warnings

High

Confidence: 97% confidence
Finding: The skill explicitly promotes large-scale automatic submission across multiple platforms with high-volume and parallelized application behavior, but it does not present a strong user-facing warning that this will perform real external actions on third-party services using candidate data. In context, the skill is designed to interact with job platforms and submit applications at scale, so missing consent and risk notices materially increases the risk of account misuse, accidental mass submissions, and policy violations.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The tracking and alerting features imply storage of application histories and transmission of updates by email, but the document does not clearly warn users that candidate/application data may be retained, processed, and sent through external channels. Because this skill targets recruiters and agencies handling multiple candidates, the omitted disclosure raises privacy, confidentiality, and compliance risks for personal data.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The API mode exposes endpoints for batch application operations and application data retrieval without clear warnings about binding scope, authentication, authorization, or handling of sensitive candidate records over the network. In this skill's context, those endpoints could expose personal data or trigger external actions remotely if deployed insecurely, making the omission operationally dangerous.

Ssd 4

Medium

Confidence: 98% confidence
Finding: The guidance explicitly frames scheduled, staggered application submission as a way to avoid being flagged as spam, which is evasion-oriented behavior rather than legitimate rate-management. In the context of a bulk auto-apply tool with exec capability, this encourages abuse of third-party platforms and increases the likelihood of deceptive automation at scale.

Ssd 2

Medium

Confidence: 95% confidence
Finding: The statement that frequency control is used to avoid restrictions encourages anti-detection or anti-rate-limit behavior instead of straightforward compliant usage. Given the skill's purpose of mass job applications, such phrasing normalizes platform-abuse tactics and can facilitate automation designed to bypass service controls.

Ssd 2

Medium

Confidence: 97% confidence
Finding: Presenting account rotation as a troubleshooting measure implies using alternate identities or accounts to continue automated activity after friction or restrictions, which is a classic abuse-evasion pattern. In a job-application automation context, this materially increases the risk of account misuse, policy evasion, and deceptive bulk submissions across platforms.

Static analysis

No suspicious patterns detected.