Back to skill

Security audit

Think-Plan

Security checks for vulnerabilities and agentic risk

Overview

This is a planning-focused skill with disclosed workspace plan saving and optional subagent execution, with no evidence of hidden code, exfiltration, or system persistence.

Install this only if you want a Chinese-language planning workflow that asks probing questions, saves finalized plans into the workspace, and may coordinate subagents for complex execution after confirmation. Review generated plans before allowing execution, especially on broad or ambiguous tasks.

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
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (11)

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The manifest presents the skill as a planning and requirements-analysis tool, but the body expands its authority into writing files and initiating execution. This mismatch can mislead users or orchestrators into granting the skill broader operational effects than expected, increasing the risk of unintended workspace modification or task execution.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill is described entirely as operating in Chinese and presents no option for the user to choose another language. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale restriction is explicitly justified.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

Broad trigger phrases like everyday requests to 'help plan' or 'analyze a scheme' can cause unintended activation in normal conversations. Because the skill can later write files and initiate execution-related actions, overbroad activation increases the chance of surprising side effects and inappropriate delegation.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill adds multi-agent session spawning even though its stated purpose is planning, which materially increases capability and attack surface. Spawning subagents can amplify unsafe actions, propagate sensitive context, and make execution harder to audit or control if triggered under a seemingly benign planning workflow.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The low-complexity examples and keywords are broad enough to match many ordinary user requests, which can cause this planning skill to activate in situations where it is not needed. In an agent system, over-triggering can unnecessarily route benign requests into a deeper planning workflow, increasing prompt surface area, cost, latency, and the chance that downstream instructions are applied inappropriately.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The medium-complexity keywords such as general requests to analyze, research, compare, or design are ambiguous and can match a wide range of normal conversations. This creates unreliable routing boundaries, making the skill prone to accidental invocation and enabling unnecessary decomposition or multi-agent behavior for tasks that do not warrant it.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The high-complexity trigger phrases include generic wording like building a system, platform, architecture, or project from scratch, which may over-classify many requests as highly complex. In context, that can push the agent into oversized planning or layered-agent orchestration, amplifying operational risk and reducing predictability without user consent.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

This markdown file contains user-facing operational instructions solely in Chinese, including headings, prompts, and workflow guidance. Under the policy rule, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale limitation is explicitly justified, which is not present here.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The red-line rule states the skill must provide at least 2 and at most 3 options, but the later '方案数量' section says '严格限制 1-3 个方案'. This creates an intent-level contradiction in the skill documentation about core behavior.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
77% confidence
Finding

All instructional content, examples, and trigger phrases are presented exclusively in Chinese, with no indication that language selection is optional or region-specific. If this file is used as operational guidance for a general-purpose skill, it may impose a locale preference without user opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

This markdown file contains natural-language guidance exclusively in Chinese, which can amount to a language-policy issue if skills are expected to avoid forcing a specific language without user opt-in. There is no indication that the content is intended only for a Chinese-speaking audience or that users can select another language.

Content

No source excerpt is available for this finding.