Security audit

Security Ib Compliance

Security checks for vulnerabilities and agentic risk

Overview

This is a static investment banking compliance reference skill with no executable behavior, though some sample AML/KYC code and regulatory claims should be verified before use.

Install only as a reference aid. Do not rely on its static regulatory summaries or sample code for final compliance decisions without checking official sources and qualified counsel, especially for AML/KYC and current securities rules.

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
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (3)

Ae4

Medium

Category: analysis-evasion
Confidence: 80% confidence
Finding: Suspicious Unicode normalization or mixed-script content

Intent-Code Divergence

Medium

Confidence: 91% confidence
Finding: The docstring at L367-L369 presents this as a working AML/KYC review routine, but at L386-L389 the code tries to read `results["sanction_screening"]` and `results["transaction_monitoring"]` while `results` is still being constructed. That is an active contradiction between the documented intent and the actual code behavior, because the function would fail instead of completing the advertised review.

Intent-Code Divergence

Medium

Confidence: 85% confidence
Finding: At L395 the function is documented as performing sanctions-list screening, but the return block at L399-L404 is misindented relative to the `if similarity > 0.85:` at L398. As written, the code does not reliably implement the documented screening logic, creating an intent-code divergence rather than a mere omission.

Static analysis

No suspicious patterns detected.