Back to skill

Security audit

CFD Model Develop

Security checks for vulnerabilities and agentic risk

Overview

The skill is coherent for FNO model training, but it tells the agent to install packages, download mutable remote model code, and later run that downloaded code without clear user approval or integrity controls.

Install only if you are comfortable with the agent contacting ModelScope, installing Python packages, caching downloaded code, and executing that code locally. Use an isolated environment, review or pin the ModelScope package and model revision, and avoid exposing credentials or sensitive datasets unless you trust the upstream source.

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Findings (1)

T03 · Remote Payload Retrieval and Execution

Error
Location
SKILL.md:35
Finding

Unpinned Remote Dependencies and Mutable Model Code Are Downloaded and Executed

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill instructs the agent to automatically install a Python package and download remote model code as the mandatory first action, before obtaining explicit user confirmation for environment changes or network access. This is dangerous because it can modify the runtime, pull unreviewed third-party code into the environment, and create unintended outbound connections, which is especially risky in enterprise or restricted environments.

Content

No source excerpt is available for this finding.