academic-suite-v1

• Outcome: Review
• Latest audit: May 28, 2026, 4:36 AM
• Version:
  1.0.0

Overview

This academic writing suite is broadly coherent, but it deserves Review because it force-installs many unpinned agent skills and includes a mandatory de-AI rewriting stage for academic work.

Install only if you trust this publisher and are comfortable with the suite updating multiple OpenClaw skills. Review the listed dependencies first, avoid --force unless you intend to replace existing skills, provide only narrow API credentials or knowledge-base paths, and independently verify citations, facts, and your institution or publisher's AI-disclosure rules before using generated work.

Publisher note

Complete Academic Research & Writing Suite — One-click installation of the full Academic Pipeline ecosystem with all 8 dependencies

SkillSpector

By NVIDIA

Vulnerability Patterns

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

Findings (8)

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: The README instructs users to run an installation script and force-install commands without explaining what files will be created, modified, or overwritten. This can lead users to execute a script that changes their environment or pulls additional code without informed consent, increasing the risk of unintended system modification or supply-chain exposure.

Missing User Warnings

Medium

• Confidence: 84% confidence
• Finding: The literature-search stage explicitly uses external platforms and APIs, but the README does not warn that user prompts, research topics, metadata, or document contents may be transmitted to third-party services. In an academic workflow, this can expose unpublished research ideas, sensitive source material, or institutional data to external providers without the user's awareness.

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: The skill advertises one-click installation of a multi-component academic workflow that performs networked search, API-backed retrieval, and possible access to IMA knowledge-base data, but it does not clearly warn users what data may be transmitted to external services. In a research-writing context, prompts, drafts, references, and repository paths may contain sensitive or proprietary information, so lack of disclosure meaningfully increases privacy and data-handling risk.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The mandatory HUMANIZE/去 AI 化 stage is described as altering manuscript text, but the documentation does not adequately warn that this may materially change meaning, citations, tone, or authorship-relevant characteristics. In academic settings, undisclosed transformation of final text can create integrity, attribution, and compliance problems even if framed as stylistic editing.

Natural-Language Policy Violations

High

• Confidence: 98% confidence
• Finding: Mandating a 去 AI 化 / HUMANIZE stage directs the system to transform output to conceal or minimize AI-generated characteristics, which can facilitate evasion of AI-use disclosure requirements, review scrutiny, or policy controls. In an academic-paper pipeline, this is especially dangerous because it is tightly coupled to submission-ready workflows, increasing the likelihood of deceptive use in coursework, peer review, or publication contexts.

Vague Triggers

Medium

• Confidence: 79% confidence
• Finding: The manifest description advertises a broad, one-click "complete" academic suite without stating clear trigger boundaries, task scope, or user-confirmation requirements. In a skill ecosystem, vague activation language can cause over-broad invocation and increase the chance that high-impact subordinate skills or install flows are run in contexts the user did not explicitly intend, especially given this package aggregates 8 dependencies and an install script.

Ssd 4

Medium

• Confidence: 96% confidence
• Finding: A mandatory '去 AI 化处理' stage is explicitly designed to disguise AI-generated academic writing before finalization. In the context of academic submission, this materially increases the risk of deliberate policy evasion, academic misconduct, and deceptive authorship representation.

Ssd 4

Medium

• Confidence: 95% confidence
• Finding: The suite advertises dedicated English and Chinese '去 AI 化' capabilities as bundled features, showing that concealment of AI authorship is a core product function rather than an incidental wording issue. In an academic context, that makes the skill more dangerous because it operationalizes evasion across the full research and writing pipeline.

VirusTotal

VirusTotal findings are pending for this skill version.