Back to skill

Security audit

日报生成器专业版

Security checks for vulnerabilities and agentic risk

Overview

This daily-report skill is not clearly malicious, but it asks for broad file, command, API, credential, and external callback capabilities without enough scoping or user-control detail.

Install only if you are comfortable granting this skill broad agent authority. Before use, restrict it to specific repositories, calendars, task systems, output folders, and approved webhook domains; avoid enabling Bash or callbacks unless needed, and do not provide broad API credentials without clear connector-level limits.

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

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest markets a narrow daily-report generator, but the body expands the skill into broad automation involving file access, external integrations, and command execution. This scope mismatch can cause users or an agent runtime to grant more trust and permissions than the declared purpose warrants, increasing the chance of unsafe tool use.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

Advertising command execution for a daily-report skill is dangerous because it normalizes use of Bash for a task that does not inherently require shell access. In an agent environment, this can lead to unnecessary execution of system commands, expanding attack surface to command injection, filesystem modification, or environment data exposure.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill instructs use of API keys and external service connections without clearly constraining what services are accessed or what data may be transmitted. For a daily-report tool, this broad integration language can prompt over-privileged configuration and unintended disclosure of internal work data to third-party systems.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

An overly broad natural-language trigger can cause the skill to activate in contexts where the user did not intend to invoke it. Because this skill also advertises file, API, and command-related capabilities, accidental invocation could lead to unnecessary data processing or external/system actions.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

The example trigger phrase is so generic that it can match ordinary conversation, making unintentional activation more likely. In an agent setting, generic triggers are risky when paired with tools that can read, write, call networks, or execute commands.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill states it can automatically collect content from Git, calendars, and task systems, but it does not prominently warn users that potentially sensitive workplace data may be accessed and aggregated. This can lead to silent over-collection of internal activity data beyond user expectations.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

A callback URL feature creates a clear exfiltration path because generated reports, metadata, or logs may be sent to arbitrary external endpoints. Without a warning or restrictions, users may unknowingly transmit sensitive work information outside trusted boundaries.

Content

No source excerpt is available for this finding.