Security audit

日志脱敏工具

Security checks for vulnerabilities and agentic risk

Overview

The skill is a local log-sanitization guide with expected sensitive-log access, but users should manage the unredacted backup files it creates.

Use preview mode first, scan only intended log paths, avoid relying on callback_url unless the publisher documents exactly what is sent, and delete or protect .bak backups because they contain the original sensitive log data.

A.I.G

Vulnerability Patterns

Insecure DependenciesIntroduces malicious components through unsafe dependency sources
Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
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

Findings (2)

T09 · Insecure Skill Coding Practices

Warning

Location: SKILL.md:177
Finding: Plaintext Backups Preserve Sensitive Log Data After Redaction
Content: View full analysis
Remediation: View remediation

T08 · Insecure Dependencies

Note

Location: SKILL.md:332
Finding: Unpinned Third-Party Dependency Installation
Content: View full analysis
Remediation: View remediation

SkillSpector

Vulnerability Patterns

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
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Intent-Code Divergence

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: The documentation promises 'local only' and 'zero data exfiltration', yet the declared input schema includes a callback_url for asynchronous notification. Even if not implemented, documenting a remote callback path creates ambiguity that can lead users or downstream agents to transmit scan results or metadata to external endpoints, undermining the privacy guarantees of a log-sanitization tool.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The trigger scope is described with a very broad set of keywords and scenarios, without firm boundaries or exclusions. In an agent context, overbroad activation can cause the skill to run on unintended files, logs, or workflows, increasing the chance of unnecessary access to sensitive data and accidental modification or disclosure of unrelated content.
Content: No source excerpt is available for this finding.

Intent-Code Divergence

Low

Category: Not specified by scanner
Confidence: 83% confidence
Finding: The 'known limitations' section says the free version does not support 'MCP工具流水线集成', suggesting no agent/tooling integration. Elsewhere, the file states the skill runs on AI Agent platforms, depends on an 'LLM API' supplied by the agent platform, and is an 'MD+EXEC' skill driven by agent-executed commands, which conflicts with the stated unsupported integration boundary.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 81% confidence
Finding: L351 明确写出“完全中文化表达，适配中文开发者工作流”，全文也未提供语言切换、用户偏好选择或适用范围说明。对于通用技能说明，这种单一语言约束可能与要求尊重用户语言/地区偏好的组织策略相冲突。
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.