Back to skill

Security audit

日历

Security checks for vulnerabilities and agentic risk

Overview

This calendar skill is not clearly malicious, but its ability to create and delete Google Calendar events is under-scoped and its documentation is inconsistent.

Review this before installing. Only use it where the agent must ask before creating or deleting events, and verify the actual gcalcli and Google OAuth setup yourself because the dependency and credential documentation is unreliable.

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
Vulnerability Patterns
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (6)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill advertises deletion of calendar events but does not prominently warn that it can remove or modify user data. In the context of a calendar-management skill with write and exec capabilities, missing destructive-action disclosure increases the risk of accidental data loss, especially if an agent invokes the skill automatically.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The invocation guidance is broad and encourages use across project management, planning, tracking, team collaboration, and automation workflows without precise trigger boundaries. In a skill capable of deletion and other state-changing actions, vague activation criteria increase the likelihood of accidental invocation and unintended modification of user calendar data.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The documented return schema is unrelated to calendar operations and instead describes a grading/evaluation report. This mismatch can mislead an agent or user about what the skill actually does, increasing the risk of unsafe automation, incorrect downstream parsing, or concealment of real side effects such as calendar modification or deletion.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill claims to manage Google Calendar via gcalcli, but the dependency section lists only an LLM API and omits the actual calendar tooling. This inconsistency can cause agents to invoke the skill under false assumptions, potentially routing data to the wrong service or attempting privileged actions without the declared controls or dependencies actually present.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The API key configuration references a generic API_KEY environment variable, which conflicts with the claimed Google Calendar/gcalcli integration model. This can confuse users into exposing unrelated credentials or encourage unsafe secret handling practices while obscuring the true authentication mechanism used for calendar access.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

The description states '支持中文交互' as a built-in behavior, which can be interpreted as imposing a language mode rather than offering a selectable locale. The file does not clearly say the user may choose their preferred language or that multilingual behavior is optional.

Content

No source excerpt is available for this finding.