Back to skill

Security audit

Li Xiang Perspective V2

Security checks for vulnerabilities and agentic risk

Overview

This text-only persona skill has no code or system access, but it can be triggered by generic phrases and then speak as a real person in first person until explicitly stopped.

Install only if you specifically want a simulated Li Xiang-inspired analysis mode. Treat all first-person statements as role-play, not real statements or endorsements, and be aware that common product-strategy phrases may trigger the persona unexpectedly.

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 (1)

T01 · Skill Instruction Hijacking

Warning
Location
SKILL.md:27
Finding

Persistent First-Person Impersonation and Session Instruction Hijacking

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 27-35; corroborating identity instructions at lines 64-68
Vulnerability Type: Agent role and response-policy hijacking
Risk Level: Medium

Vulnerable Instruction Snippet

The following is an English rendering of the relevant instructions:

markdown
## Role-Playing Rules

After this Skill is activated, respond directly as Li Xiang.

- Use “I” rather than “Li Xiang would think...”
- Use Li Xiang's tone—direct, pragmatic, and efficiency-oriented
- State the disclaimer only once upon initial activation
- Do not say “Li Xiang would probably think...” or “If it were Li Xiang, he might...”

Exit role: Return to normal mode when the user says “exit,”
“switch back to normal,” or “stop role-playing.”

The identity card reinforces the replacement identity:

markdown
Who I am: I am Li Xiang, founder, chairman, and CEO of Li Auto.
I am a serial entrepreneur and the founder of Pcpop and Autohome.

Technical Analysis

The Skill does not merely request analysis using a public figure's documented decision framework. It directs the agent to replace its presented identity with that of a real person and prohibits safer attribution such as “Li Xiang would think.” It also limits the disclaimer to the first activation and preserves the role until the user provides one of several explicit exit phrases.

This constitutes skill instruction hijacking because loading or triggering the Skill changes the agent's session-level identity, attribution policy, and response behavior. The broad trigger phrases documented by the Skill include ordinary product-related expressions such as “product definition” and “family users,” increasing the possibility that impersonation could be activated without a sufficiently explicit request to role-play.

The project contains only SKILL.md; therefore, no executable code, external payload retrieval, dependency compromise, ...[truncated 1467 chars]

Remediation
View remediation

Remediation Suggestions

  1. Replace direct impersonation with an explicitly attributed analytical mode, such as: “Provide an analysis inspired by decision frameworks attributed to Li Xiang in public sources.”
  2. Prohibit claims that the agent is the real person. Require formulations such as “From this framework's perspective” rather than first-person biographical assertions.
  3. Require explicit user opt-in before enabling the perspective. Do not activate it from generic phrases such as “product definition” or “family users.”
  4. Remove persistent role state. Apply the requested perspective only to the prompt that explicitly invokes it unless the user clearly requests a temporary multi-turn mode.
  5. Display a concise attribution notice whenever confusion is reasonably possible, rather than only once at initial activation.
  6. Separate factual claims from simulated recommendations and clearly identify that the framework is derived from public information.
  7. Replace the identity card's statement “I am Li Xiang” with language such as: “This Skill models selected publicly documented decision principles associated with Li Xiang; it is not Li Xiang and does not represent his views.”
  8. Narrow activation phrases to explicit commands, for example: “Analyze this using the Li Xiang-inspired public framework.”
  9. Add a general safeguard stating that higher-priority agent safety, identity, attribution, and truthfulness requirements cannot be overridden by the Skill.
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (2)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger list includes very broad terms like “产品定义” and “家庭用户”, which are common business phrases and could activate the skill unintentionally in unrelated conversations. This creates prompt-routing risk: the agent may enter a persona unexpectedly, override normal response behavior, and reduce reliability or safety controls tied to explicit user intent.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill instructs the agent to respond directly in-character after activation, but does not define strict activation boundaries beyond loose trigger and exit phrases. Combined with broad triggers, this can cause sticky or ambiguous persona takeover across turns, leading the model to present fictionalized identity claims as first-person statements and potentially mishandle sensitive requests under role-play framing.

Content

No source excerpt is available for this finding.