Security audit

Scrapling Boss Recruitment

Security checks for vulnerabilities and agentic risk

Overview

This recruiting automation skill is purpose-aligned, but it asks for live account cookies and can collect candidate personal data, send resumes to an LLM, and perform bulk outreach without enough safeguards.

Install only if you control the Boss account, have authorization to process candidate data, and are comfortable providing sensitive session cookies. Keep cookie and proxy credentials out of chats, logs, and shared files; avoid bulk greetings or resume collection without legal/privacy review and clear rate limits; and disable or constrain external LLM parsing unless candidates and your organization have approved that data flow.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (8)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 90% confidence
Finding: The skill clearly enables network-driven scraping of Boss直聘 and operational actions like candidate search, resume retrieval, and batch greetings, yet it declares no permissions. This mismatch weakens governance and user awareness because the agent may perform external requests and data collection without explicit capability disclosure or approval boundaries.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill facilitates bulk messaging and resume collection, which involves personal data processing and potentially unsolicited outreach, but the warning section only mentions account ownership, robots.txt, and cookie validity. Without clear privacy, consent, retention, and lawful-use safeguards, the skill can be used to harvest candidate data or spam users at scale.

Missing User Warnings

High

Confidence: 97% confidence
Finding: The document explicitly instructs users to extract and provide live Boss直聘 authentication cookies, including core session tokens such as __zp_stoken__. These are effectively bearer credentials: anyone who obtains them may be able to impersonate the user session, access recruiting data, and perform actions on the account. In a scraping/recruitment automation skill, this is more dangerous because the workflow normalizes credential harvesting and reuse without clear warnings, scoping, or secure handling guidance.

Missing User Warnings

Medium

Confidence: 89% confidence
Finding: The proxy example embeds username and password directly in the proxy URL, which encourages insecure secret handling and increases the chance credentials are copied into source files, logs, screenshots, or error messages. While presented as an example, the lack of warning or safer alternatives can lead users to hardcode reusable proxy credentials in plaintext. In this skill context, that risk is amplified because automation tooling often logs configuration and is frequently shared across operators.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The file documents selectors and dynamic-fetching steps for scraping candidate profiles, including personal and employment data such as name, age, gender, education, work history, and self-description, but provides no privacy, consent, retention, or lawful-use guidance. In a recruitment scraping skill, this omission materially increases the risk of unauthorized collection and misuse of personal data, especially when combined with automation features like search, bulk outreach, and resume extraction.

Missing User Warnings

Medium

Confidence: 89% confidence
Finding: The code retrieves and structures candidate resume data, including profile and employment history, using authenticated scraping without any built-in consent, authorization, or privacy controls. In a recruitment-scraping skill, this materially increases privacy and compliance risk because it enables bulk access to personal data that may be restricted by platform terms, user expectations, or applicable data-protection laws.

Missing User Warnings

Medium

Confidence: 82% confidence
Finding: The class accepts raw authenticated cookies and reuses them for privileged network access, but provides no safeguards for secure storage, redaction, scope limitation, or validation. In this skill's context, stolen or mishandled cookies could grant account-level access to candidate search and resume data, amplifying both account compromise and privacy exposure.

Missing User Warnings

Medium

Confidence: 98% confidence
Finding: The code sends raw resume text to an LLM client, and resumes commonly contain highly sensitive personal data such as phone numbers, email addresses, age, education, salary, and employment history. In a recruitment scraping skill, this materially increases privacy and compliance risk because candidate PII may be transferred to external model providers without consent, minimization, or disclosure.

Static analysis

No suspicious patterns detected.