Back to skill

Security audit

Research Paper Writer

Security checks for vulnerabilities and agentic risk

Overview

This skill is a low-risk academic writing helper with formatting references and no evidence of hidden data access, persistence, or harmful actions.

Install only if you want an academic paper writing helper or compatibility alias. Be aware that the legacy trigger may catch generic writing prompts, and the runtime implementation appears minimal, so functionality may depend on the surrounding skill system rather than this package alone.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Memory PoisoningPersistent Context Injection, Context Window Stuffing, Memory Manipulation
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (9)

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/acm_formatting_specs.md (reported line 1)May include surrounding context.

md
# Academic Writing Style Guide

This guide extracts writing conventions from high-quality academic papers on context-aware systems and large vision-language models.

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/ieee_formatting_specs.md (reported line 1)May include surrounding context.

md
# Academic Writing Style Guide

This guide extracts writing conventions from high-quality academic papers on context-aware systems and large vision-language models.

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · references/writing_style_guide.md (reported line 1)May include surrounding context.

md
# Academic Writing Style Guide

This guide extracts writing conventions from high-quality academic papers on context-aware systems and large vision-language models.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The legacy trigger phrase write introduction 振动控制 is broad enough to collide with ordinary user writing requests, which can cause the wrong skill to activate unintentionally. In an agent environment, ambiguous invocation can misroute prompts, bypass a user's intended workflow, or produce actions under the wrong skill context, making prompt-routing behavior less predictable and easier to abuse.

Content

No source excerpt is available for this finding.

Context Window Stuffing

Medium
Category
Memory Poisoning
Confidence
80% confidence
Finding

Skill attempts to fill the context window with filler content, displacing legitimate instructions and safety constraints. This can degrade agent performance or bypass safety boundaries.

Content

Scanner excerpt · references/acm_formatting_specs.md (reported line 311)May include surrounding context.

md
Table 1: Accuracy Results

Dataset     Baseline    Proposed
鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€
Dataset A      87.3%      94.2%
Dataset B      89.1%      95.7%
Dataset C      85.7%      93.4%

Context Window Stuffing

Medium
Category
Memory Poisoning
Confidence
80% confidence
Finding

Skill attempts to fill the context window with filler content, displacing legitimate instructions and safety constraints. This can degrade agent performance or bypass safety boundaries.

Content

Scanner excerpt · references/acm_formatting_specs.md (reported line 315)May include surrounding context.

md
Table 1: Accuracy Results

Dataset     Baseline    Proposed
鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€
Dataset A      87.3%      94.2%
Dataset B      89.1%      95.7%
Dataset C      85.7%      93.4%

Context Window Stuffing

Medium
Category
Memory Poisoning
Confidence
80% confidence
Finding

Skill attempts to fill the context window with filler content, displacing legitimate instructions and safety constraints. This can degrade agent performance or bypass safety boundaries.

Content

Scanner excerpt · references/ieee_formatting_specs.md (reported line 226)May include surrounding context.

md
PERFORMANCE COMPARISON

Method          Accuracy    Time (ms)
鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€
Baseline        87.3%       145
Proposed        94.2%       132
鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€

Context Window Stuffing

Medium
Category
Memory Poisoning
Confidence
80% confidence
Finding

Skill attempts to fill the context window with filler content, displacing legitimate instructions and safety constraints. This can degrade agent performance or bypass safety boundaries.

Content

Scanner excerpt · references/ieee_formatting_specs.md (reported line 229)May include surrounding context.

md
PERFORMANCE COMPARISON

Method          Accuracy    Time (ms)
鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€
Baseline        87.3%       145
Proposed        94.2%       132
鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€鈹€

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

The file presents key user-facing instructions in Chinese while later sections are in English, but it does not state whether language is user-selectable or why this bilingual presentation is required. This can amount to an implicit language policy decision without user opt-in.

Content

No source excerpt is available for this finding.