Back to skill

Security audit

历史论述题教练

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed Chinese K12 history essay coaching skill with optional, consent-gated learning profile and reminder features.

Install only for the intended Chinese K12 history-learning context. Before enabling profile memory, cross-skill sharing, or reminders for a student, confirm the student or guardian understands what will be saved, shared, and how to pause or delete it.

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 (10)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The manifest description is entirely written as Chinese-facing instructions for students and trigger phrases, with no indication that users may choose another language or locale. Under the stated policy, a skill that effectively constrains interaction to a specific language should offer user opt-in or clearly document a justified locale restriction.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
82% confidence
Finding

The file is entirely written as a mandatory protocol in Chinese and includes required teacher-facing labels such as 【AI 生成,入库前请人工验算】 without offering a language choice or stating that the skill is restricted to a Chinese-language educational context. Under the policy, natural-language instructions that force a specific language or locale without opt-in can be a violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

This JSON schema embeds user-facing descriptions and enumerated values entirely in Chinese, including required value sets such as consent actors and grade bands. Because the file does not document that the skill is region-specific or provide any user opt-in for language/locale, it imposes a specific language context in a way that can violate language/locale policy.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

Natural-language descriptions throughout the schema are written entirely in Chinese, including the title/description and operational notes, with no indication that language choice is optional or that the skill is intentionally limited to a Chinese-only locale. This creates a language-policy concern because the file embeds a fixed language requirement without documenting user opt-in or a justified region-specific constraint.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

The manifest describes a history essay coach that guides students in forming and defending their own historical arguments, explicitly limiting itself to checking and coaching rather than broader learner-record operations. This distributed schema explicitly includes wrong_answer_handover, subject_profile_writeback, and reminder_enqueue, enabling the skill to transfer error records, write back to a learning DNA/profile system, and enqueue reminders—capabilities that go beyond the narrow tutoring/checking role described in the manifest.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

This schema permits the history essay coach to send subject_profile_writeback data that mutates learner-profile fields under subjectExtensions.history. Even though the writable area is narrowed, it still creates a cross-skill profile mutation channel that may exceed the coach's stated 'only check and guide' role and could cause silent persistence of inferred student traits or performance signals without clear necessity or minimization.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The schema allows this skill to enqueue reminders with arbitrary content, scheduling, priority, and optional non-mergeable delivery. That creates a cross-skill action channel outside the coach's narrow educational checking role, increasing the risk of unwanted messaging, consent misuse, or behavioral nudging if the skill over-collects or overuses learner interaction data.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This file contains user-facing instructional policy text only in Chinese, and it does not indicate that users may choose another language or locale. Under the language/locale policy rule, forcing a specific language without user opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

L11 明确说明本库按“中国大陆的中文 K12 场景”设计,并将示例话术、安全求助渠道默认设为中国大陆。这构成了默认语言/地区约束;虽然 L13-L17 说明跨地区部署前需本地化,但并未向最终用户提供语言或地区选择/确认机制。

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
74% confidence
Finding

This line requires historical evaluation to follow '课程标准和统编教材为准', which imposes a specific national curriculum and locale framing. A locale-specific constraint can be acceptable, but the file does not clearly present this as a justified region-specific limitation or opt-in choice.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.