ClawHub
Back to skill

Security audit

OSTEO-GC: Glucocorticoid-Induced Osteoporosis Modeling

Security checks for vulnerabilities and agentic risk

Overview

The skill is a self-contained osteoporosis modeling script with no network, persistence, credential, or hidden execution behavior, but users should treat its treatment outputs as clinical decision support rather than medical advice.

Install only if you understand it is an educational or clinician-support calculation tool. Its osteoporosis risk and treatment recommendations should be reviewed by a licensed clinician and should not be used alone to start, stop, taper, or choose medication.

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
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The code prints patient-specific fracture risk categories and concrete treatment recommendations such as bisphosphonates, denosumab, teriparatide, monitoring intervals, and glucocorticoid tapering guidance without any warning that the output is a simplified educational model rather than validated medical advice. In a clinical context, users may reasonably rely on this output for real treatment decisions, which can lead to harmful self-management or inappropriate care because the model uses approximations and omits many clinical contraindications and patient-specific factors.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.