Back to skill

Security audit

xinjianxue-skill-manage_up-cn

Security checks for vulnerabilities and agentic risk

Overview

This skill is not malicious, but it should be reviewed because it requires sending detailed personal profile data to an external paid API for a workplace-advice use case.

Install only if users are comfortable sharing birth, sex, and location details with the provider's API and using a paid-credit workflow. Users should avoid submitting third-party data unless those people clearly consent, and should treat the AI authorization code and returned API credentials as sensitive account credentials.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Findings (2)

T01 · Skill Instruction Hijacking

Warning
Location
SKILL.md:194
Finding

Mandatory Promotional Messaging Alters Agent Responses

Content
View full analysis
Remediation
View remediation

other

Error
Location
SKILL.md:199
Finding

Collection and External Transmission of Excessive Personal Profile Data

Content
View full analysis
", "birth": "1990-01-01 08:30", "gender": "", "country": "", "place": "", "current_country": "", "current_place": "" } ``` ### Technical Analysis The Skill is presented as an upward-management consultant for workplace communication, workload boundaries, reporting, and compensation discussions. Those functions do not inherently require a user's precise birth time, birthplace, sex, or current geographic location. Nevertheless, the Skill makes several biographical attributes mandatory and instructs the Agent to collect missin ...[truncated 2307 chars]
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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