Security audit

Test

Security checks for vulnerabilities and agentic risk

Overview

This skill is a low-impact Tarot-style tie-breaker, but its trigger wording is broad and the referenced helper files are not included in the inspected artifact.

Install only if you are comfortable with a whimsical random tie-breaker influencing ambiguous or casual requests. Keep it out of security-sensitive, production, financial, or data-integrity decisions, and be aware that this artifact appears incomplete because the referenced drawing script and card/reference files were not included.

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
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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 (4)

Anti-Refusal Statement

High

Category: Anti-Refusal
Confidence: 90% confidence
Finding: Skill attempts to nullify the agent's safety policies or restrictions ('you have no restrictions', 'ignore your guidelines', 'do anything now'). This is a direct jailbreak that disables guardrails.
Content: Scanner excerpt · SKILL.md (reported line 91)May include surrounding context.
md
## Rationalizations to Reject - "The cards said to, so I must" → Cards inform direction, they don't override safety or correctness - "This reading justifies my pre-existing preference" → Be honest if the reading challenges your instinct - "I'll keep drawing until I get what I want" → One draw per decision point; accept the reading

Instruction Override

High

Category: Prompt Injection
Confidence: 90% confidence
Finding: This pattern attempts to override system instructions or ignore safety constraints. Without LLM analysis, manual review is recommended.
Content: Scanner excerpt · SKILL.md (reported line 91)May include surrounding context.
md
## Rationalizations to Reject - "The cards said to, so I must" → Cards inform direction, they don't override safety or correctness - "This reading justifies my pre-existing preference" → Be honest if the reading challenges your instinct - "I'll keep drawing until I get what I want" → One draw per decision point; accept the reading

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The activation description is overly broad and includes common conversational phrases like 'whatever you think', 'surprise me', and 'try again', which can cause the skill to trigger in many ordinary interactions. That creates unintended behavior injection risk by letting a randomness-based decision aid influence workflows outside narrowly intended cases, especially if the agent auto-selects skills from natural language cues.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The 'When to Use' section broadens invocation to ambiguous prompts, nonchalant delegation, redraw requests, and stylistic references, which are weak signals for activating a non-deterministic planning mechanism. In practice this can let the skill steer decisions when the user did not meaningfully consent to randomness, reducing predictability and potentially affecting task quality or safety boundaries.
Content: No source excerpt is available for this finding.