ClawHub
Back to skill

Security audit

Financial Engineer Digital Employee

Security checks for vulnerabilities and agentic risk

Overview

This is a Markdown-only financial modeling reference skill that does not bundle executable code, though users should treat any datasets and generated reports/models from its examples as sensitive.

Install only as a reference skill. If you adapt and run the example commands, use authorized and minimized datasets, redact PII, review all reports and model artifacts before saving or sharing them, and do not treat the outputs as financial, legal, insurance, or automated decision advice without qualified human review.

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 (2)

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill declares itself as advisory-only with no execution, persistence, or storage, yet the body repeatedly instructs running local scripts that ingest datasets, train models, and write reports, models, logs, and JSON artifacts to disk. This mismatch can mislead users, reviewers, or policy gates into granting the skill more trust than warranted, causing sensitive financial or personal data to be processed and persisted contrary to the stated safety posture.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The text claims the skill will not automatically read or save business data or PII, but the documented modules require loading dataset files and producing saved outputs derived from them. In a financial modeling context, those outputs may contain sensitive aggregates, sample previews, feature names, scores, or model artifacts that encode regulated data, so the inaccurate claim can create unsafe assumptions about data handling.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.