Security audit

Competitor Tracker

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed competitor influencer-marketing analysis workflow with some persistent memory and prospecting-adjacent outputs users should understand before use.

Install this only if you are comfortable with competitor analysis results, opportunity lists, and some creator/competitor facts being saved into the agent's memory and proposed to a creator registry. Review or disable memory writes if the analysis includes confidential strategy, spend estimates, CRM context, or sensitive partner information.

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

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
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (6)

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The manifest explicitly says the skill is not for finding the user's own new creators, but the workflow later includes identifying untapped or former-competitor creators as opportunities for the user's brand. That is a scope expansion and can cause the agent to perform prospecting behavior the user or host may have intended to reserve for a different skill, weakening policy boundaries and increasing the chance of unauthorized data use or unexpected actions.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 90% confidence
Finding: The skill is presented as competitor analysis, but the contract expands into CRM checks and creator-registry updates that affect internal prospecting records. This broadens the operational scope from analysis into modification of internal systems, which can lead to unintended data persistence, contamination of internal records, or actions outside the user's expectation.
Content: No source excerpt is available for this finding.

Intent-Code Divergence

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The documentation says the skill is not for finding the user's own new creators, yet the later instructions direct it to rank untapped and former-competitor creators as opportunities. This contradiction can bypass skill-routing and user-consent expectations, causing the agent to take on prospecting functions under the label of competitor tracking.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The skill declares writes to persistent memory paths and promotion of facts to shared memory without a clear user-facing warning at execution time about persistence. In agent environments, silent persistence can expose sensitive business intelligence, create retention surprises, and make later tasks rely on stored competitive data the user did not realize would be saved.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium

Category: Not specified by scanner
Confidence: 97% confidence
Finding: The template explicitly includes 'Untapped Influencers' and 'Former Competitor Partners' sections that guide the agent to identify available creators for the user's own program. That conflicts with the stated skill boundary ('Not for finding your own new creators') and can cause the agent to perform out-of-scope prospecting under the guise of competitor analysis. In this context, the danger is policy/scope drift and misuse of the skill for lead-generation rather than a classic exploit, but the instructions materially expand what the agent is encouraged to do.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Low

Category: Not specified by scanner
Confidence: 86% confidence
Finding: The manifest describes this as a competitor-tracking skill that benchmarks the user's program, but the template asks for a full side-by-side table containing detailed fields for '[Your Brand]' such as active partners, spend, engagement rate, content style, and relationship type. That goes beyond pure competitor tracking and requires substantive analysis of the user's own program, which is broader than the rest of the described behavior.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.