Security audit

chat

Security checks for vulnerabilities and agentic risk

Overview

This chat-style skill is not clearly malicious, but it asks for broad file, command, and API authority that is not well scoped for its stated purpose.

Install only if you intentionally want a broadly empowered automation skill, not just a chat-style helper. Restrict read/write/exec permissions where possible, require explicit approval before commands, file writes, API calls, or bulk messaging, and avoid giving it sensitive files, API keys, or private communication records until retention and access controls are clarified.

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

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands

Findings (7)

Description-Behavior Mismatch

High

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The skill is presented as a communication-preference/chat helper, but the documentation also declares broad file handling, external API access, and system command execution as core functions. This creates a dangerous capability mismatch: an agent or user may invoke a seemingly low-risk chat skill that can actually access files, call external services, or execute commands, increasing the chance of unauthorized data access or host-side impact.
Content: No source excerpt is available for this finding.

Context-Inappropriate Capability

High

Category: Not specified by scanner
Confidence: 98% confidence
Finding: System command execution is not justified by the stated purpose of learning communication preferences or adapting response style. Exposing exec in a chat-oriented skill materially increases the risk of prompt-driven command execution, local data exposure, environment reconnaissance, or chaining into more serious compromise if the host environment is permissive.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 87% confidence
Finding: The description repeatedly says 'Use when 需要营销推广、广告投放、获客转化、增长裂变时使用', which is a very broad invocation condition covering many common business tasks rather than a narrowly scoped trigger. The file does not provide explicit trigger phrases, exclusion rules beyond one illegal-marketing note, or concrete negative examples to clarify when the skill should and should not activate.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 82% confidence
Finding: The description says '支持中文交互，无需复杂配置即开即用', and other sections emphasize Chinese interaction as a built-in behavior. This can be read as imposing a language/locale expectation without clearly offering the user a language choice at the point of activation or description.
Content: No source excerpt is available for this finding.

Intent-Code Divergence

Medium

Category: Not specified by scanner
Confidence: 84% confidence
Finding: The documentation claims strong privacy protection while elsewhere describing archiving and retrieval of communication records. This contradiction can mislead users about data retention and handling, causing oversharing of sensitive content under false assumptions about how the skill stores or processes communications.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: The skill advertises file writing, API calls, and command execution without prominent user-facing warnings about possible system and data impact. In an agent setting, insufficient disclosure increases the risk that users approve or invoke the skill without understanding it can alter files, transmit data externally, or run host commands.
Content: No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The skill advertises generic file read/parse/write behavior even though its stated function is chat/style adaptation. Unnecessary file access broadens the attack surface for sensitive data exposure, unintended file modification, or misuse by an agent that assumes the skill is only handling conversational preferences.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.