Security audit

Resilience Building Architect

Security checks for vulnerabilities and agentic risk

Overview

This skill is a self-help resilience planner with no network access, persistence, credential use, or hidden system behavior, though its mental-health examples would benefit from clearer crisis-escalation guidance.

Use this as reflective resilience support only. For serious distress, self-harm thoughts, or feeling that life has no meaning, users should contact local emergency services, a crisis hotline, or a qualified mental-health professional instead of relying on the skill alone.

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

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
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (4)

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The Chinese-language scenario includes a user statement that 'every day I wake up and feel life has no meaning,' which is a potential suicidality or acute mental health crisis signal. The expected output offers recovery coaching and peer-group suggestions but does not explicitly direct the user to immediate crisis, emergency, or licensed mental health support, creating a risk of under-triaging a high-risk user in a mental-health-adjacent skill.

Natural-Language Policy Violations

Low

Confidence: 83% confidence
Finding: The skill description presents example interactions in both Chinese and English, but it does not state how language is selected or whether users can choose their preferred language. Under the policy rule, forcing or implicitly assuming a language/locale without opt-in can be a natural-language policy issue.

Natural-Language Policy Violations

Low

Confidence: 76% confidence
Finding: The document is otherwise written in English but includes a full scenario and expected output in Chinese, without explaining that multilingual responses are optional or user-selected. This can conflict with a language/locale policy requiring explicit user opt-in rather than silently switching languages.

Natural-Language Policy Violations

Low

Confidence: 91% confidence
Finding: The manifest explicitly sets the skill language to "en", which imposes a language choice without indicating that users can select another language. Under the policy, locale or language constraints should either be optional for the user or clearly justified as region-specific.

Static analysis

No suspicious patterns detected.