Back to skill

Security audit

Los Angeles

Security checks for vulnerabilities and agentic risk

Overview

This is a markdown-only Los Angeles guide with some practical-advice caveats, but no hidden code, persistence, credentials, or system-level behavior.

Install only if you want LA-specific guidance. Treat prices, transit details, safety conditions, park access, allergy information, and food-safety advice as starting points to verify with current official or local sources before acting.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (8)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The activation condition is very broad: 'User asks about Los Angeles for any purpose' can match a large share of unrelated or only tangentially related queries. This can cause unintended invocation, leading the agent to load this skill when a more appropriate skill should handle the request, which increases the chance of irrelevant guidance, priority conflicts, or misrouting of user intent.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The food safety section gives oversimplified heuristics such as treating a long line as evidence that street food is safe and implying servers generally know ingredients, without warning that these are not reliable protections against foodborne illness or allergen exposure. In a practical city guide, readers may treat this as actionable real-world safety advice, which can lead to illness or allergic reactions if they rely on these cues instead of verified food handling practices and direct ingredient confirmation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
94% confidence
Finding

The line 'Soju is required' uses unnecessarily prescriptive language that frames alcohol consumption as mandatory. In a travel/food guidance skill, this can pressure users who are underage, sober, pregnant, medically restricted, or otherwise unable or unwilling to drink, creating unsafe or exclusionary guidance.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
31% confidence
Finding

The content is written entirely in English and includes region-specific shorthand like "DTLA" and "Westside." However, this appears to be ordinary location guidance rather than a policy that forces a language or locale on the user, so this is weak and likely not a violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

The Glendale section ties the neighborhood identity to an ethnic community and includes subjective framing such as 'can feel insular' without context. In a location-guidance skill, this can reinforce stereotypes or make users infer social exclusion based on ethnicity, creating biased or discriminatory guidance even if no harm was intended.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
78% confidence
Finding

The title anchors the skill to a specific city, which is a locale constraint expressed in natural language. Because the document does not offer a broader location choice or explain that the content is intentionally region-specific as part of a clearly scoped regional tool, it may conflict with locale-choice policy expectations.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
96% confidence
Finding

The file is explicitly scoped to Los Angeles, as stated in the title, and all guidance is specific to that region. This is a locale constraint presented as the document's stated purpose rather than a forced language or locale policy imposed on users.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

The itinerary recommends a Joshua Tree day trip involving hiking and desert travel but provides no safety guidance about heat, water, sun exposure, cell coverage, or return-after-dark risks. In a travel-planning skill, users may rely on concise itinerary advice as actionable guidance, so omission of basic preparation warnings can contribute to preventable harm.

Content

No source excerpt is available for this finding.