Back to skill

Security audit

Biohazard Crime Scene Cleanup Kit

Security checks for vulnerabilities and agentic risk

Overview

This is a markdown-only marketing prompt kit for Nevada biohazard cleanup companies, with no hidden code or persistence, but users should verify legal and privacy-sensitive content before publishing.

Install only if you are using it for Nevada/Clark County biohazard cleanup marketing or will have local counsel/regulators revalidate the legal references. Treat all generated ads, review requests, and public responses as drafts, especially for trauma, hoarding, crime-victim, insurance, or coroner-related situations.

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
  • 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 (5)

Anti-Refusal Statement

High
Category
Anti-Refusal
Confidence
80% confidence
Finding

Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.

Content

Scanner excerpt · examples/silver-shield-biohazard-las-vegas.md (reported line 159)May include surrounding context.

md
**Compliance Audit — Page 1:**
- Gate 1: PASS — written ECP, annual training, Hep B cited
- Gate 2: PASS — NDEP permit # in credentials
- Gate 3: PASS — scene release process explained; no "complete cleanup" without caveat
- Gate 4: N/A
- Gate 5: PASS — ABRA verified, HAZWOPER; no fake credentials
- Gate 6: PASS — NRS 217 explained with DHHS contact; "we don't handle your claim" explicit

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The prompt hard-codes Nevada- and Clark-County-specific legal and procedural guidance throughout, but it does not clearly restrict use to that jurisdiction or require confirmation from the user. If reused elsewhere, the model may generate legally inaccurate instructions, credential claims, and public response language that are noncompliant in another state, creating regulatory and liability risk.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
85% confidence
Finding

The prompt's compliance guidance states that matters involving coroner/remains should be escalated to private contact immediately and not discussed publicly. However, the later response templates include public explanation of crime-scene release timing involving law enforcement and the coroner, which conflicts with the stated highest-liability handling rule.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill generates review-request templates for families, hoarding clients, crime-scene victims, and other highly sensitive customers without an explicit privacy-risk warning or consent guardrail beyond TCPA. Encouraging public reviews in these contexts can pressure vulnerable people to disclose traumatic events, create reputational harm, and expose protected or stigmatizing personal circumstances through a public platform.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The prompt says Gate 3 and Gate 6 issues should move immediately to private contact, but several templates still provide substantive public handling first. In a high-sensitivity domain involving crime scenes, insurance disputes, and possible human remains, even limited public explanations can reveal regulated or highly private facts, create legal exposure, or steer users into noncompliant public statements.

Content

No source excerpt is available for this finding.