Security audit

xingtuTaskAuthor

Security checks for vulnerabilities and agentic risk

Overview

The skill appears to perform the promised XingTu export, but it asks users to paste a live browser cookie into chat and stores it in plaintext for reuse.

Install only if you are comfortable letting the agent handle an active XingTu session cookie and saving creator/contact data locally. Prefer a version that uses a browser-mediated or secure credential flow, avoids plaintext cookie storage, lets you exclude or mask contact fields, and asks before writing task metadata to memory.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (3)

Missing User Warnings

High

Confidence: 98% confidence
Finding: The skill explicitly instructs the user to copy and paste a full browser Cookie string into chat and then persists it to `~/.xingtuCookie.txt`. Session cookies are bearer credentials; anyone with access to the chat transcript, logs, local filesystem, or memory artifacts could reuse them to impersonate the user and access the XingTu account.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The skill exports a dataset containing potentially sensitive personal and business information about authors, including identifiers, city, contact handle (`微信号`), performance metrics, and application notes, and also records processing details in memory. Without explicit minimization, consent, retention limits, or privacy warning, this creates unnecessary disclosure and persistence risk for third-party data.

Ssd 3

High

Confidence: 99% confidence
Finding: This is a direct natural-language exfiltration path: the user is told to retrieve the Cookie header from browser tools and paste it into the conversation, after which it is saved locally for reuse. That design unnecessarily routes highly sensitive authentication material through conversational channels and persistent storage, greatly increasing the chance of credential leakage or unauthorized reuse.

Static analysis

No suspicious patterns detected.