Security audit

supplier-risk-scoring

Security checks for vulnerabilities and agentic risk

Overview

This is a static vendor-risk scoring guide with no executable behavior, though users should treat its data collection and geography-based scoring as policy inputs that need legal and compliance review.

Before installing or using this skill, treat it as a decision-support checklist rather than an authoritative policy. Have procurement, legal, compliance, and privacy stakeholders approve the data sources, retention practices, and any geography-based scoring rules before they affect vendor selection, PO holds, or contract decisions.

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 (2)

Natural-Language Policy Violations

Medium

Confidence: 86% confidence
Finding: The geographic scoring rubric assigns preset risk penalties to broad countries and regions, including specific adverse treatment for countries like China and entire regional groupings, without requiring documented business justification, objective evidence, or local legal review. In practice this can produce biased or overgeneralized vendor decisions, creating discrimination, fairness, and compliance exposure if users treat the table as authoritative rather than contextual.

Missing User Warnings

Low

Confidence: 90% confidence
Finding: The skill explicitly directs users to gather potentially sensitive vendor information such as business credit reports, self-reported financials, references, legal history, and compliance records, but provides no guidance on lawful basis, minimization, retention, access control, or secure handling. In a procurement context this can lead to unnecessary collection or unsafe storage of confidential business and personal data, increasing privacy, confidentiality, and contractual risk.

Static analysis

No suspicious patterns detected.