Security audit

Vancouver

Security checks for vulnerabilities and agentic risk

Overview

This is a markdown-only Vancouver guide with some ordinary content-quality cautions, but no hidden execution, persistence, credential handling, or deceptive behavior.

Safe to install as a local reference guide. Treat prices, immigration, tax, licensing, health, and food-safety advice as starting points, not authoritative current advice; verify high-stakes details with official sources or qualified professionals.

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

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 (6)

Vague Triggers

Medium

Confidence: 96% confidence
Finding: The activation rule says to use the skill whenever the user asks about Vancouver for any purpose, which is broad enough to capture many generic conversations without clear routing boundaries. This can cause inappropriate skill invocation, irrelevant guidance, or accidental overshadowing of more specialized skills, though it does not by itself create code-execution or data-exfiltration risk.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The file recommends eating spot prawns as sashimi with only a vague note of 'extremely fresh only' and does not provide a clear food safety warning about raw seafood risks such as parasites, bacterial contamination, temperature abuse, and heightened risk for pregnant or immunocompromised people. In a consumer travel/food guide, readers may treat this as practical advice and attempt unsafe preparation or consumption without understanding the hazards.

Missing User Warnings

Low

Confidence: 90% confidence
Finding: This business guide presents specific tax rates, wage amounts, immigration pathways, and licensing requirements as if they are current, but it does not warn users that these figures and rules can change frequently. A user could rely on outdated information when incorporating, hiring, filing taxes, or pursuing immigration-related business options, leading to financial penalties, compliance issues, or poor business decisions.

Missing User Warnings

Low

Confidence: 79% confidence
Finding: This markdown file includes advice to use vitamin D supplements and obtain a light therapy lamp as ways to cope with winter conditions. Because these are health-affecting recommendations, the description would be safer if it included a brief warning that medical needs vary and users should consult a clinician for significant symptoms or before starting supplements.

Natural-Language Policy Violations

Low

Confidence: 77% confidence
Finding: The section is framed around 'ESL Programs' and private language schools, but it does not explicitly acknowledge user language preference or offer multilingual alternatives beyond a separate French section. This can be read as privileging English-language integration by default, which may conflict with a language-choice policy if the organization expects neutral or opt-in language framing.

Natural-Language Policy Violations

Low

Confidence: 88% confidence
Finding: The document is entirely scoped to Vancouver neighborhoods, transit, rents, and local safety context, which effectively forces a specific geographic locale. Under the policy rule, locale constraints should either be user-selected or clearly documented as region-specific and justified; this file does not explicitly state that limitation as a documented scope or opt-in.

Static analysis

No suspicious patterns detected.