Security audit

人性博弈顾问

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed Chinese-language interpersonal analysis skill with no hidden code or system access, but its advice style is adversarial and should be used cautiously.

Install this only if you want Chinese-language, strategy-heavy interpersonal analysis. Treat its leverage and escalation advice as perspective, not instructions to pressure or manipulate people, and adapt it to legal, ethical, and cultural boundaries.

A.I.G

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

SkillSpector

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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 (10)

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 84% confidence
Finding: The metadata and description present the skill as operating through a specific cultural and analytical framing without signaling user choice, which can silently steer outputs into a fixed worldview. This is mainly a safety and UX issue rather than a direct exploit, but in a sensitive interpersonal-advice skill it can bias advice, overfit to one cultural lens, and reduce appropriateness for users from other contexts.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The trigger conditions are broad enough to auto-activate on many ordinary conversation or relationship questions, which can cause the skill to seize control of interactions outside a clearly scoped use case. In this skill, that overreach is more concerning because the downstream behavior pushes deep motive inference and strategic interpersonal advice, increasing the chance of manipulative or unwarranted analysis being applied by default.
Content: No source excerpt is available for this finding.

Ssd 4

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The skill explicitly operationalizes interpersonal escalation: it instructs the model to diagnose leverage, identify what the other party fears, determine when 'soft' approaches are exhausted, and then escalate via pressure, third parties, formal channels, or tactics designed to turn 'your problem' into 'their problem.' In context, this is not merely neutral conflict-resolution guidance; it creates a reusable framework for coercive social engineering, harassment, and manipulation across workplace, family, intimate, and negotiation scenarios.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 90% confidence
Finding: The instruction to 'fully consider' a China-specific social lens bakes in a fixed locale-specific interpretive framework at the policy level, regardless of user background or the actual context of the scenario. In an advice skill that already makes strong inferences about motives and power, this can amplify stereotyping, miscalibration, and inappropriate strategic recommendations.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The trigger phrases are highly generic conversational utterances such as '他什么意思', '我该怎么回', and '怎么看这个人', which can easily appear in ordinary user speech outside an explicit request for this skill. That raises the risk of unintended activation, causing the skill to engage in sensitive interpersonal or psychological analysis when the user may not have intended it, increasing privacy and safety concerns in a high-manipulation subject area.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: This markdown file is written entirely in Chinese and does not provide any indication that other languages are supported or that the user can opt into the locale. Under the language/locale policy rule, a skill that effectively requires a specific language without user choice can be a policy violation unless the constraint is clearly documented and justified.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 83% confidence
Finding: The file title and all instructional content are written exclusively in Chinese, with no indication that language selection is optional or that the skill is intentionally region-specific. Under the stated policy, forcing a specific language without user opt-in or documented justification is a natural-language policy concern.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The document explicitly centers a China-specific social norm framework ('在中国社交中') as a general analysis principle without user choice, qualification, or alternatives. In a skill that claims broad applicability across workplace, family, and negotiation scenarios, this can bias outputs toward one cultural lens, causing overgeneralized or misleading advice for users from other locales or subcultures.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 78% confidence
Finding: The display name, description, author string, and triggers are entirely in Chinese, which indicates the skill is presented in a single language without any opt-in or language-selection mechanism. Under the policy, locale or language restrictions should either offer user choice or be clearly justified as region-specific.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 89% confidence
Finding: The cited passage makes a broad, locale-specific claim about Chinese society as if it were generally applicable, without framing it as context-dependent or optional. In a skill that gives strategic advice about relationships, work, and negotiation, this can bias outputs toward cultural stereotyping or overgeneralized guidance, reducing reliability and potentially causing inappropriate advice for users outside that context.
Content: No source excerpt is available for this finding.