Security audit
Security checks for vulnerabilities and agentic risk
This is a prompt-only marketing kit with disclosed Nevada-focused compliance guidance and no executable behavior, but users should verify legal claims before publishing generated ads.
Installers should use this as a Nevada-focused marketing prompt kit, not as legal advice or a substitute for attorney review. Before publishing generated ads or website copy, verify the shop’s license, AGSC status, ADAS capability, glass sourcing, insurance statements, and whether the Nevada-specific rules apply to the intended market.
The skill hardcodes Nevada-specific legal rules and licensing requirements into a generic campaign generator without clearly scoping use to Nevada or requiring user confirmation of jurisdiction. In compliance-oriented marketing content, this can cause users in other states to receive inaccurate legal guidance and publish misleading or unlawful ads.
The ADAS gate requires mandatory referral/disclosure language in every ad and campaign when the shop lacks ADAS capability, but the output section only conditionally includes ADAS content and can omit the disclosure entirely. That can produce materially misleading automotive safety advertising by implying a windshield replacement is complete without required recalibration warnings.
The skill declares a hard block against deductible-absorption style claims, but later email templates still make deductible-specific insurance statements that can be interpreted as legal/compliance advice or misleading coverage representations. In a marketing-generation skill, this contradiction is dangerous because users may publish noncompliant copy believing it was pre-vetted by the prompt.
The compliance section explicitly prohibits unsupported superlative ranking claims unless there is a documented source. The later output instruction requiring 'Why [Shop Name] Is #1 for Windshield Replacements in [City]' directly conflicts with that rule because '#1' is a ranking/superiority claim presented without any evidentiary condition.
Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.
Every auto glass shop using ChatGPT or a generic AI tool is generating content that:
- Says "safe to drive immediately" without verifying adhesive cure time (AGSC Standard 150 violation)
- Advertises "we handle everything" without disclosing ADAS recalibration — a NHTSA-flagged safety issue
- Uses "no out-of-pocket" or "we pay your deductible" — **explicitly illegal in Nevada** (NRS 686A.2825; DOI enforcement)
- Claims "OEM glass" while installing aftermarket — federal false advertising (FTC Act Section 5)
- Runs review incentives without FTC 2023 disclosure requirements
Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.
Every auto glass shop using ChatGPT or a generic AI tool is generating content that:
- Says "safe to drive immediately" without verifying adhesive cure time (AGSC Standard 150 violation)
- Advertises "we handle everything" without disclosing ADAS recalibration — a NHTSA-flagged safety issue
- Uses "no out-of-pocket" or "we pay your deductible" — **explicitly illegal in Nevada** (NRS 686A.2825; DOI enforcement)
- Claims "OEM glass" while installing aftermarket — federal false advertising (FTC Act Section 5)
- Runs review incentives without FTC 2023 disclosure requirements
Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.
Every auto glass shop using ChatGPT or a generic AI tool is generating content that:
- Says "safe to drive immediately" without verifying adhesive cure time (AGSC Standard 150 violation)
- Advertises "we handle everything" without disclosing ADAS recalibration — a NHTSA-flagged safety issue
- Uses "no out-of-pocket" or "we pay your deductible" — **explicitly illegal in Nevada** (NRS 686A.2825; DOI enforcement)
- Claims "OEM glass" while installing aftermarket — federal false advertising (FTC Act Section 5)
- Runs review incentives without FTC 2023 disclosure requirements
Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.
Every auto glass shop using ChatGPT or a generic AI tool is generating content that:
- Says "safe to drive immediately" without verifying adhesive cure time (AGSC Standard 150 violation)
- Advertises "we handle everything" without disclosing ADAS recalibration — a NHTSA-flagged safety issue
- Uses "no out-of-pocket" or "we pay your deductible" — **explicitly illegal in Nevada** (NRS 686A.2825; DOI enforcement)
- Claims "OEM glass" while installing aftermarket — federal false advertising (FTC Act Section 5)
- Runs review incentives without FTC 2023 disclosure requirements
Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.
Every auto glass shop using ChatGPT or a generic AI tool is generating content that:
- Says "safe to drive immediately" without verifying adhesive cure time (AGSC Standard 150 violation)
- Advertises "we handle everything" without disclosing ADAS recalibration — a NHTSA-flagged safety issue
- Uses "no out-of-pocket" or "we pay your deductible" — **explicitly illegal in Nevada** (NRS 686A.2825; DOI enforcement)
- Claims "OEM glass" while installing aftermarket — federal false advertising (FTC Act Section 5)
- Runs review incentives without FTC 2023 disclosure requirements