English Reading Coach

• Outcome: Pass
• Latest audit: Sep 16, 2026 · 8:47 AM UTC
• Version: 1.0.0

Overview

This is a straightforward English reading-coach skill with minor routing and language-preference caveats, not a security threat.

Install this if you want an English reading tutor workflow. Be aware it may activate on broad English-reading prompts, and scheduled drill framing may switch to Vietnamese when the agent thinks it is helpful; set your preferred language and ask for confirmation if that matters.

A.I.G

Vulnerability Patterns

• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally
• Unauthorized Access and Privilege Escalation: Obtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (8)

Vague Triggers

High

• Confidence: 97% confidence
• Finding: The skill advertises activation on very broad, casual phrases such as generic requests to read or practice English. Overbroad triggers can cause inappropriate auto-invocation, leading the agent to enter a long instructional workflow when the user did not clearly request that behavior and increasing the chance of context hijacking or irrelevant tool use.

Vague Triggers

High

• Confidence: 98% confidence
• Finding: The trigger description is excessively broad and includes casual phrases such as general requests to read in English, which can cause the skill to activate outside clear user intent. Overbroad invocation creates routing and consent problems: users may be pushed into a lengthy reading workflow, web fetching, or exam-style coaching when they did not explicitly ask for this specialized behavior.

Natural-Language Policy Violations

Medium

• Confidence: 85% confidence
• Finding: The README consistently defines the skill as an English reading coach and all example interactions are English-specific, but it does not state that the user can choose another language or explicitly opt into English-only operation. Under the stated policy, forcing a specific language without opt-in is a natural-language policy concern.

Vague Triggers

Medium

• Confidence: 94% confidence
• Finding: The core scope repeats ambiguous trigger language without constraints, reinforcing that the skill should activate for loosely related requests. This increases unintended activation risk across normal conversation and can override more appropriate skills or default assistant behavior.

Natural-Language Policy Violations

Medium

• Confidence: 87% confidence
• Finding: The scheduled drill instructs the agent to keep framing in Vietnamese 'when helpful' without confirming the user's language preference. This can violate locale and user-preference expectations, and in some environments may cause inappropriate language switching or policy noncompliance for users who did not request Vietnamese output.

Whitespace Padding

Medium

• Category: Prompt Injection

Content

Before the learner reads the full text, assign a **pre-reading task** matched to the target strategy:

| Strategy         | Pre-reading task                                                                                                  |
| ---------------- | ----------------------------------------------------------------------------------------------------------------- |
| Skimming         | "Read only the title, first sentence, and last sentence. What do you think the text is about?"                    |
| Scanning         | "I'll give you 3 questions. Find the answers as fast as you can — don't read every word."                         |

• Confidence: 70% confidence
• Finding: Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.

Natural-Language Policy Violations

Medium

• Confidence: 95% confidence
• Finding: The file is explicitly framed as an "English Reading Coach" and repeatedly instructs the use of "simple English" and English-only reading activities, but it does not indicate that this language constraint is optional or limited to a justified region-specific use case. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation.

Natural-Language Policy Violations

Low

• Confidence: 92% confidence
• Finding: The file is explicitly framed as an 'English Reading Coach' and instructs use of English reading content, but it does not indicate that the user can choose another language or explicitly opt into an English-only experience. Under the policy, forcing a specific language without user opt-in is a natural-language locale violation.

Static analysis

No suspicious patterns detected.