ClawHub
Back to skill

Security audit

报告汇总构建器基础版

Security checks for vulnerabilities and agentic risk

Overview

The skill is a report-summarization helper, but its privacy and credential-handling instructions are too inconsistent for sensitive work reports without review.

Review before installing. Use it only with report directories you are comfortable exposing to the active agent and any configured LLM/API provider. Avoid running the broad environment-variable check, and confirm output paths or backups before allowing it to write generated reports.

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (5)

Description-Behavior Mismatch

High
Confidence
95% confidence
Finding
The skill markets itself as local-only and focused on report summarization, yet it also supports callback URLs, external API use, and network-dependent behavior elsewhere in the document. This mismatch can mislead users into providing sensitive work-report data under false assumptions, increasing the risk of unintended exfiltration to remote services.

Intent-Code Divergence

High
Confidence
97% confidence
Finding
The document explicitly claims FREE-version data is stored locally and not uploaded to the cloud, but other sections require network connectivity and reference external APIs. For a tool processing potentially sensitive internal reports, contradictory privacy claims create a serious trust and data-leakage risk because users may not realize their content could be sent off-host.

Context-Inappropriate Capability

Medium
Confidence
90% confidence
Finding
The skill instructs users to enumerate environment variables matching API, KEY, TOKEN, or SECRET, which is unrelated to core report-building functionality and encourages sensitive secret discovery. Even with masking in the example, such guidance normalizes secret inspection and may expose the presence, naming, and management of credentials to the agent or logs.

Vague Triggers

Medium
Confidence
84% confidence
Finding
The trigger condition is overly broad, covering generic analysis, reporting, statistics, and visualization scenarios. This can cause the skill to activate in unrelated contexts and perform file reads/writes or other actions on business data without clear user intent, increasing the chance of unintended data handling.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The skill describes outputting generated reports directly to report directories but does not clearly warn users or require confirmation before modifying files. In an agent context with write and exec permissions, this can lead to silent overwrites, unintended persistence, or corruption of work artifacts.

Static analysis

No suspicious patterns detected.