Security audit

IELTS Speaking Coach

Security checks for vulnerabilities and agentic risk

Overview

This IELTS coaching skill is coherent and purpose-aligned, with voice/audio privacy and broad activation behavior worth noting before use.

Install only if you are comfortable sending IELTS answers, and possibly voice recordings, through the assistant's ASR and LLM workflow. Use text-only mode for sensitive material, exit IELTS mode when finished, and confirm whether any optional backend in your environment stores transcripts or learning history.

A.I.G

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

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (15)

Ae1

High

Category: analysis-evasion
Confidence: 100% confidence
Finding: Referenced artifact was not completely inspected
Content: Scanner excerpt · SKILL.md (reported line 127)May include surrounding context.
md
- `vocab-map.json` — Topic-aware vocabulary upgrades

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: Line L44 describes standardized scoring and correction outputs with fixed Chinese templates. Forcing a specific language in user-facing behavior, without noting opt-in or a locale-specific justification, is a natural-language policy concern under the language/locale rule.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The README explicitly encourages users to send voice messages, convert them with ffmpeg, and transcribe them with ASR, but it does not clearly disclose how speech data is processed, whether audio/transcripts leave the device, or what retention/sharing rules apply. Because spoken responses may contain personally identifying or sensitive information, the lack of a clear privacy warning and consent notice creates a real data-handling risk.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 84% confidence
Finding: SQP-3 applies to all file types and covers language or locale policy violations. The README presents Chinese commands such as '模拟考试', '学习路线', and several fixed entry/menu phrases as activation examples, but does not explicitly offer a language-selection option or clarify that these are optional examples alongside equivalent alternatives for all functions.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 88% confidence
Finding: The file recommends and models specifically British English usage, such as "university," "flavourful," and IELTS-style phrasing, without stating that this is optional or tied to a chosen locale. Because the content is instructional natural language, this can amount to a locale preference being imposed rather than offered as a user choice.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The file is entirely written in Chinese and provides no indication that users can choose another language or that Chinese is a required locale. In a broadly accessible IELTS coaching skill, this can exclude or mislead users who expect English or multilingual support, causing accessibility and usability failures rather than code execution risk.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 86% confidence
Finding: The file title and description are entirely in Chinese and specify guidance for Chinese native speakers, but there is no indication that users can opt into another language or that the skill is intentionally restricted to a Chinese-only audience. Under the language/locale policy, forcing a specific language without user opt-in is a natural-language policy concern.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The activation rule says to trigger on 'any IELTS speaking request' and remain active until an explicit exit command, which creates unclear boundaries and sticky behavior. In practice, this can hijack subsequent user turns, override user intent in mixed conversations, and increase the chance of unintended shell/network-capable skill execution once activated.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 98% confidence
Finding: Forcing Chinese output for scoring responses without offering language choice can mislead or exclude users who are operating in another language, especially in a multilingual assistant environment. While not a direct code-execution issue, it is a real safety and usability concern because it can cause misunderstanding of assessment results and corrective advice.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The trigger list is unusually broad and includes generic phrases such as 'correct my grammar', 'model answer', 'study plan', and common Chinese learning phrases that could match ordinary tutoring or conversation requests outside a clearly scoped IELTS context. This can cause the skill to activate unexpectedly, take over unrelated interactions, and apply its scoring or mode-persistence behavior when the user did not intend to invoke this skill.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 76% confidence
Finding: The file includes activation and menu phrases in Chinese alongside one English trigger, but does not state whether the skill supports multiple languages or requires Chinese by design. This can indicate a language/locale constraint without clear user opt-in, especially for commands like "学习路线" and "模拟考试" elsewhere in the guide.
Content: No source excerpt is available for this finding.

Missing User Warnings

Low

Category: Not specified by scanner
Confidence: 87% confidence
Finding: The description advertises audio input, pronunciation scoring, and ASR processing but provides no notice about how voice data is handled, stored, transmitted, or whether third-party tooling/services are involved. For a skill that processes spoken user data, this omission can mislead users about privacy expectations and increases the risk of improper consent or unsafe deployment in environments with sensitive audio.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 81% confidence
Finding: This markdown file presents the skill content in Chinese/English and cites Chinese-language sources, but it does not indicate that the bilingual or Chinese-oriented format is optional, user-selected, or restricted to a justified locale-specific audience. Under the policy, fixed language or locale behavior without user opt-in can be a natural-language policy issue.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 92% confidence
Finding: This markdown file includes bilingual headings and labels such as the Chinese question-bank label and section title alongside English content. Under the stated policy, forcing or assuming a specific language/locale without offering user choice can be a natural-language policy concern.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Low

Category: Not specified by scanner
Confidence: 82% confidence
Finding: The description and vocabulary entries prescribe registers such as "spoken|neutral|formal|academic" and include locale-specific terms like "flavourful", "chuffed", and "recognise", indicating a fixed English variant and style assumption. The file does not offer any user opt-in or alternative locale/language handling, which can violate language/locale choice policy.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.