Back to skill

Security audit

agentscope-benchmark-runner

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward benchmark-file generator whose local file writes are disclosed, user-directed, and aligned with its purpose.

Install only if you are comfortable running a local Python generator that creates files in a directory you choose. Use a dedicated benchmark folder to avoid overwriting similarly named generated files, and review benchmark results before sharing because prompts may appear in outputs.

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
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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 (1)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
87% confidence
Finding

The skill instructs users to run local scripts that generate benchmark files and write runner/config/result artifacts, but the manifest does not declare any tool scope such as permissions or allowed-tools. This creates a trust and containment gap: consumers cannot tell from the skill metadata that file read/write behavior is expected, which can lead to unintended filesystem access when the skill is executed in an agent environment.

Content

No source excerpt is available for this finding.