ClawHub
Back to skill

Security audit

Paper Card Analyzer

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward paper-summary helper that reads parsed paper files and writes clearly named paper-card outputs in the same folder.

Before installing, be aware that using this skill can create or overwrite paper-card.md and paper-card.json in the selected paper folder and append revision notes to paper-card-feedback.md. Keep backups or rename existing files if those outputs already contain work you want to preserve.

A.I.G

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

SkillSpector

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Rogue AgentSelf-Modification, Session Persistence
  • 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 (3)

Self-Modification

High
Category
Rogue Agent
Content
- `paper-card.md` and `paper-card.json`.
6. Request human feedback and revise:
   - Ask what to correct, expand, or make stricter.
   - Update card and save again (overwrite current files).
   - Append each round to `paper-card-feedback.md` with: round number, user request, key edits.
7. Repeat revision rounds until the user explicitly confirms satisfaction.
8. Keep uncertainty explicit:
Confidence
90% confidence
Finding
The revision workflow explicitly instructs the agent to overwrite existing output files on each round. While framed as normal editing behavior, this is still a file-integrity risk because repeated automated overwrites can destroy prior results, erase useful provenance, or clobber unexpected files if path selection is mistaken.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The skill directs the agent to automatically create and save multiple files in the local folder, and later overwrite them during revisions, without requiring an explicit confirmation step immediately before modification. In an agent environment with filesystem access, this can cause unintended changes to user data, especially if the wrong paper pair or folder is selected.

Natural-Language Policy Violations

Low
Confidence
89% confidence
Finding
The instruction says to produce the paper card in English by default, which imposes a language preference in the skill's natural-language behavior. The file does not explicitly offer the user a language choice or describe this as an opt-in locale constraint.

Static analysis

No suspicious patterns detected.