ClawHub
Back to skill

Security audit

Paper Deep Reading

Security checks across malware telemetry and agentic risk

Overview

This skill is a coherent research-paper analysis helper with local text-processing scripts, though its dependency pins and registry tags should be cleaned up.

Install this for paper-reading workflows only if you are comfortable giving the agent access to the specific paper/source folders you provide. Use a narrow input directory and local output path, verify any online sources the agent retrieves, and pin or audit markdown and pyyaml versions in controlled environments. The publisher should correct the unrelated crypto and purchase capability tags.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Known Vulnerable Dependency: markdown — 2 advisory(ies): CVE-2025-69534 (Python-Markdown has an Uncaught Exception); CVE-2025-69534 (Python-Markdown version 3.8 contain a vulnerability where malformed HTML-like se)

High
Category
Supply Chain
Confidence
77% confidence
Finding
`markdown>=3.6` permits installation of Markdown 3.8, and the finding reports known advisories affecting that version. Because this skill processes research papers and related source material that may include attacker-controlled markup, a vulnerable Markdown parser can increase risk of denial of service or parser-triggered failures when handling malformed content.

Known Vulnerable Dependency: pyyaml — 8 advisory(ies): CVE-2019-20477 (Deserialization of Untrusted Data in PyYAML); CVE-2020-1747 (Improper Input Validation in PyYAML); CVE-2020-14343 (Improper Input Validation in PyYAML) +5 more

Critical
Category
Supply Chain
Confidence
96% confidence
Finding
`pyyaml>=6.0` may allow vulnerable PyYAML releases, and PyYAML has a history of serious deserialization issues when untrusted YAML is parsed with unsafe loaders. In the context of an agent skill that may ingest paper metadata, configs, appendices, or auxiliary files from external sources, this becomes especially dangerous because untrusted YAML content is plausible and could lead to code execution or other severe compromise if loaded unsafely elsewhere in the skill.

VirusTotal

56/56 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.