Security audit

Sr

Security checks for vulnerabilities and agentic risk

Overview

This is a research-instruction skill that uses external sources but does not contain code, persistence, credential handling, or hidden behavior.

Install this as a broad research helper, and avoid including secrets, private business data, or regulated information in research prompts because the intended workflow may use external search and source sites.

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

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (4)

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The phrase "handles any research task from quick lookups to deep academic analysis" is an extremely broad invocation scope for a markdown skill description. It does not clearly constrain when the skill should activate versus when a general assistant response would be more appropriate, increasing the risk of unintended invocation collisions.

Vague Triggers

Medium

Confidence: 93% confidence
Finding: The description "Ultimate research framework combining best of top-rated skills" is overly broad and lacks clear functional boundaries, which can cause the skill to be invoked in contexts far beyond its intended purpose. In agent systems, ambiguous invocation metadata can expand the skill's effective authority and increase the chance it is selected for sensitive or unrelated tasks, creating security and safety risk through misrouting rather than direct code execution.

Vague Triggers

Medium

Confidence: 93% confidence
Finding: The skill advertises that it can handle essentially any research task and decide its own approach, which is overly broad activation language for a capability that may trigger external information gathering. Broad, underspecified scope increases the chance the agent is invoked in contexts the user did not intend, leading to unnecessary web access, data exposure in queries, or overreach into higher-risk research workflows.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The skill lists multiple third-party and web-based source types but does not warn users that prompts or derived search terms may be sent to external services. This can cause unintentional disclosure of sensitive or proprietary information and may expose users to untrusted content without informed consent.