Security audit

问题总结

Security checks for vulnerabilities and agentic risk

Overview

This skill only formats the current conversation into a user-copyable summary, with no code execution, network sending, credential access, or persistence.

Before copying the generated summary into another AI service, review it and remove secrets, credentials, private messages, internal documents, personal data, or confidential business details.

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
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (6)

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The README describes summarizing the current conversation for copying into other AI platforms but does not clearly warn users that their conversation content will be reformatted for external sharing. In this context, the core function is data transfer preparation, so missing disclosure meaningfully increases the chance that users will export sensitive or confidential information without informed consent.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 87% confidence
Finding: All user-facing instructions and examples are written as Chinese-only behavior, including explicit Chinese trigger phrases, with no indication that other languages are supported or that Chinese is an optional preference. Under the language/locale policy, a skill should not impose a specific language without opt-in or a clearly documented justification.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The trigger phrases like “总结一下问题” and “整理一下去问豆包” are broad, natural conversational language and can overlap with ordinary user requests. This can cause the skill to activate unintentionally and repackage conversation content for export to another AI platform without the user explicitly intending to invoke the skill, increasing the risk of unintended disclosure of sensitive context.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The manifest description and the entire output template are written as mandatory Chinese-language behavior, but the file does not state that this is a China-specific skill or provide any opt-in for other languages. That creates a natural-language policy concern because it imposes a specific language/locale by default.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The trigger conditions are broad and include a catch-all phrase for 'other similar expressions', which can cause the skill to activate when the user did not explicitly intend to invoke it. This creates control-flow ambiguity: the assistant may switch into summary-only behavior at the wrong time, potentially suppressing normal assistance or reshaping sensitive user content for transfer to another platform without clear consent.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The manifest description is written to define the skill as automatically outputting a structured summary in Chinese, and the metadata also tags the skill as "chinese". This creates a language-specific constraint without any indication that the user can choose another language or opt in, which matches the locale/language policy concern.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.