Back to skill

Security audit

Serial Dilution Calculator

Security checks for vulnerabilities and agentic risk

Overview

This is a simple local serial-dilution calculator with no evidence of network access, credential use, persistence, or hidden behavior, though its scientific inputs and documentation are incomplete.

Before using this for wet-lab work, verify the generated dilution series manually because the implementation ignores the documented target/final concentration. The script appears safe to run locally from a security perspective, but the skill should better document any intended file outputs if later versions add them.

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 Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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 (2)

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

The skill documentation explicitly states that local Python/R scripts are executed and that input/output files are read and written, but it does not clearly warn the user that the skill can modify the local workspace. Even without obvious malicious behavior, this is a real safety issue because users may invoke the skill without understanding that it performs local actions and persists data.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
83% confidence
Finding

The module and class documentation present the tool as generating dilution protocols, implying inputs relevant to assay targets, but calculate_dilutions accepts target_conc and never uses it. In main, the function is even called with 0 for that parameter, showing the documented intent and implemented behavior are not aligned.

Content

No source excerpt is available for this finding.