Back to skill

Security audit

healthcare-dev

Security checks for vulnerabilities and agentic risk

Overview

This healthcare coding skill is domain-relevant, but its production-like examples handle patient data, recordings, billing, and clinical records with important safeguards missing or overstated.

Install only if you want a healthcare-development reference skill and will treat its code as illustrative, not production-ready. Review and harden any generated healthcare code with qualified security, clinical, legal, and compliance experts before it touches real patient data, telemedicine recordings, billing, or operational systems.

Vulnerability Patterns
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (4)

T09 · Insecure Skill Coding Practices

Error
Location
references/examples.md:1191
Finding

DICOM data is persisted before de-identification

Content
View full analysis
Remediation
View remediation

T05 · Unauthorized Access and Privilege Escalation

Error
Location
references/examples.md:1291
Finding

Telemedicine consultation operations lack participant authorization

Content
View full analysis
{ // Update consultation status await db.query( 'UPDATE video_consultations SET status = $1, started_at = NOW() WHERE id = $2', ['in_progress', consultationId] ); // Start recording if required for documentation if (process.env.RECORD_CONSULTATIONS === 'true') { await this.startRecording(consultationId); } // Log for audit auditLogger.info('Video consultation started', { consultationId, userId, timestamp: new Date().toISOString(), }); } async endConsultation( consultationId: string, userId: string, notes?: string ): Promise { // Update consultation status await db.query( `UPDATE video_consultations SET status = $1, ended_at = NOW(), clinical_notes = $2 WHERE id = $3`, ['completed', notes, consultationId] ); // Stop recording await this.stopRecording(consultationId); // Generate consultation summary await this.generateConsultationSummary(consultationId); // Create billing record await this.createBillingRecord(consultationId); } ``` ### Technical Analysis Both methods accept a `userId`, but neither verifies that the user is authenticated, is the assigned patient or provider, or has an administrative role authorized to manage the consultation. The database updates are constrained only by `consultationId`. This is a missing object-level authorization vulnerability. Logging the supplied `userId` does not establish authorization because the value is not tied to an authenticated identity and is not checked against the consultation record. The methods also ...[truncated 1393 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Warning
Location
references/examples.md:974
Finding

Raw HL7 message content may disclose PHI through application logs

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Note
Location
references/examples.md:296
Finding

Password hashes are compared using non-constant-time string equality

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (8)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 50)May include surrounding context.

md
> 📎 **Code example 1** (typescript) — see [references/examples.md](references/examples.md)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 107)May include surrounding context.

md
> 📎 **Code example 1** (typescript) — see [references/examples.md](references/examples.md)

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The example explicitly claims HIPAA-compliant comprehensive audit logging, but several clinically significant flows mutate records without corresponding immutable audit events. In a healthcare context, missing audit coverage undermines breach investigation, non-repudiation, and regulatory accountability, and can let unauthorized changes or access go undetected.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

This markdown file embeds code that reads sensitive environment variables such as JWT and database secrets and configures database connectivity, but the surrounding markdown provides no warning that the example handles credentials and protected health information. Under the markdown-file criteria, the description should disclose behaviors that affect privacy or system integrity when presenting code that accesses secrets and persists sensitive data.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/examples.md (reported line 62)May include surrounding context.

md
},
    fhir: {
        version: 'R4',
        baseUrl: process.env.FHIR_BASE_URL || 'https://api.healthcare.org/fhir',
        supportedFormats: ['application/fhir+json', 'application/fhir+xml'],
    },
    security: {

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The example writes audit activity to a local file and also logs to the console, which can expose sensitive metadata about PHI access through process logs, terminals, aggregators, or misconfigured collectors. In healthcare systems, even access metadata can be sensitive and local file/console sinks are often insufficiently protected or retained.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

The manifest describes a healthcare technology specialist focused on standards, compliance, EHR/EMR systems, device integration, and healthcare analytics/AI. This section goes further by creating and managing video consultations, recording sessions, sending notifications, generating consultation summaries, and creating billing records, which are product-level operational capabilities not clearly implied by the stated scope.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The code conditionally records telemedicine consultations based only on an environment flag, without any demonstrated consent capture, jurisdiction checks, participant notification, or retention safeguards. In a telehealth context this can lead to unlawful recording of highly sensitive patient interactions and large-scale privacy exposure.

Content

No source excerpt is available for this finding.