Security audit

Sensory Awareness Enhancer

Security checks for vulnerabilities and agentic risk

Overview

This skill is a low-authority sensory-awareness guide with some health-adjacent wording and quality issues, but no hidden persistence, data access, network use, or destructive behavior was found.

Before installing, treat this as a reflective mindfulness-style aid, not medical or mental-health support. Avoid relying on it for trauma recovery, persistent anxiety, dissociation, or unexplained physical symptoms without qualified professional care. Expect simple JSON coaching output rather than a robust sensory-awareness program.

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
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (7)

Vague Triggers

Medium

Confidence: 87% confidence
Finding: The activation guidance includes broad triggers such as feeling disconnected, wanting groundedness, or having experienced trauma, which can steer users toward a self-help skill in situations that may require clinical screening or professional support. In this context, the danger is not code execution but unsafe scope expansion: vulnerable users may rely on the skill for trauma-related or health-adjacent concerns beyond its stated competence.

Description-Behavior Mismatch

Medium

Confidence: 96% confidence
Finding: The manifest describes a sensory-awareness skill focused on present-moment and somatic development. In code, the skill only counts words, checks for generic keywords like goal/challenge/urgent/important, and produces boilerplate recommendations such as timelines and milestone planning, which are not specific implementations of sensory or somatic awareness.

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The trigger phrase "present moment" is highly generic and overlaps with normal conversation, making accidental or unintended activation more likely. In an agent ecosystem, broad triggers can cause the wrong skill to engage unexpectedly, leading to confusing behavior, prompt hijacking opportunities through misrouting, or unauthorized execution of the skill in contexts where it was not intended.

Natural-Language Policy Violations

Low

Confidence: 83% confidence
Finding: The skill description and scenarios alternate between Chinese and English, which implies language-specific behavior but does not state how the assistant should choose a language for users. Under the policy rule, forcing or assuming a language without opt-in can be a natural-language policy issue.

Natural-Language Policy Violations

Low

Confidence: 81% confidence
Finding: The file is primarily written in English but includes a fully Chinese scenario and expected output, with no statement explaining when Chinese responses should be used or whether language choice is based on user preference. This can create a locale-policy ambiguity because the skill appears to force or assume a language in at least one scenario without explicit opt-in guidance.

Description-Behavior Mismatch

Low

Confidence: 88% confidence
Finding: Fallback content like 'Consult professionals if needed,' 'Schedule regular reviews,' and generic 'Implementation checklist' frameworks broadens the skill into unspecified life-coaching territory. That behavior is not justified by the manifest's narrower purpose of enhancing sensory awareness and somatic intelligence.

Natural-Language Policy Violations

Low

Confidence: 88% confidence
Finding: The manifest sets the language to "en" without indicating that users can opt into another language or that the English-only constraint is required for a documented reason. This can be a language policy issue when no user choice or justification is provided.

Static analysis

No suspicious patterns detected.