Back to skill

Security audit

hermes-clean-probe

Security checks for vulnerabilities and agentic risk

Overview

The skill is marked suspicious because scanners report that its script can read a caller-selected file beyond the single reference file described as its purpose.

Consider whether the skill needs to read files beyond its bundled sample. Install only if you are comfortable with that broader read capability and the runtime limits on file access; those limits were not confirmed by the scan.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (3)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The stated single-file behavior is narrower than the implementation, which accepts a caller-selected input path without restricting it to the bundled reference. Without runtime filesystem enforcement, this permits out-of-scope reads; disclosure is constrained by JSON parsing and sum-only output, and no network exfiltration or permission-enforcement bypass is demonstrated.

Content

No source excerpt is available for this finding.

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: file_read.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

L08–L10 allow reading a caller-selected file outside the single declared reference file.

Medium
Category
Excessive Agency
Confidence
97% confidence
Finding

The script passes argv[1] directly to open(), allowing absolute paths or traversal outside the declared reference file. If runtime permissions are not enforced, a caller can read and derive output from other compatible JSON files or trigger errors and resource consumption; the script does not directly print arbitrary file contents.

Content

No source excerpt is available for this finding.