Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md | 文档与理论文本 | 本 `SKILL.md`、`README.md`、其中「理论依据」段与一切理论表述 | **不在 MIT 覆盖范围内**：保留所有权利（All rights reserved） |
Security audit
code-mentor
Security checks for vulnerabilities and agentic risk
Overview
This is a coherent programming-mentor skill with some broad activation and licensing/privacy notes, but no evidence of hidden execution, exfiltration, persistence, or malicious behavior.
Install only if you want a programming mentor that may read, search, and write project files when helping with code. Review the broad trigger phrases and the bundled learning log before use, and treat the license as mixed: the manifest says MIT, while the skill text reserves rights over documentation/theory content.
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
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
- Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (7)
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The trigger keywords are extremely generic ('code', 'mentor', '学编程'), which can cause the skill to activate for broad, unrelated user requests. In an agent ecosystem, overbroad activation can lead to unintended invocation, prompt/context hijacking opportunities, or routing users into this skill when another, more appropriate skill should handle the task.
- Content
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The manifest advertises many broad trigger phrases such as 'teach me', 'debug this', 'code review', and general learning/help requests that can match a wide range of ordinary conversations. This increases the chance of unintended invocation or over-selection of the skill, causing prompt-routing mistakes, privacy leakage of user code or text into the wrong skill, and reduced user control over which capability is activated.
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 82% confidence
- Finding
The manifest description is entirely written in Chinese and presents the skill as a general-purpose programming mentor, but it does not state that language is user-selectable or that the skill is intentionally limited to Chinese-speaking users. For a general skill, this can create an implicit language/locale constraint without user opt-in.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 81% confidence
- Finding
Natural-language policy checks apply to all file types. The leading licensing notice is written primarily in Chinese with only limited English glosses, which may impose a language expectation on users without stating that a language choice is available or why this locale constraint is necessary.
- Content
Vague Triggers
Low
- Category
- Not specified by scanner
- Confidence
- 78% confidence
- Finding
This is a markdown file, so vague-trigger review applies. The scope notice describes licensing boundaries for different file types, but it does not provide concrete operational constraints or exclusions beyond broad references to other files, which can leave readers unclear about exactly what content is covered and when.
- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 77% confidence
- Finding
The note on L15 states the learner prefers using the Chinese phrase "喂料" to request learning content, and the document overall is written primarily in Chinese. For a policy finding, this can indicate a language preference being assumed rather than explicitly offered as a choice, especially if the skill uses this preference by default in future interactions.
- Content