Security audit

Can Bus Toolkit

Security checks for vulnerabilities and agentic risk

Overview

This markdown-only skill is not clearly malicious, but it should be reviewed because it describes broad automated data handling, external alerts, persistent indexing, and rollback with unclear safeguards.

Install only if you intend to use it for controlled provenance/audit workflows. Before enabling it, restrict which logs or traffic it may inspect, disable adapters you do not need, review webhook and OTS destinations, choose report/snapshot directories deliberately, and require manual approval before any rollback or restoration action.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Intent-Code Divergence

Medium

Confidence: 90% confidence
Finding: The documented input/output schema describes a generic scoring or audit skill rather than the CAN-bus provenance, timestamping, indexing, and reporting behavior advertised elsewhere. This mismatch can cause an agent or user to invoke the skill under false assumptions, leading to inappropriate data handling, incorrect automation, or unsafe downstream decisions based on the wrong contract.

Intent-Code Divergence

Medium

Confidence: 84% confidence
Finding: The skill claims it is not suitable for real-time stream processing, yet later advertises HTTP adapters, message-queue integration, event-driven triggers, and webhook alerting that are inherently real-time or near-real-time capabilities. Contradictory operational claims can mislead deployment decisions and cause the skill to be used in environments with incorrect safety, performance, or monitoring assumptions.

Vague Triggers

Medium

Confidence: 79% confidence
Finding: The activation language is overly broad, telling agents to use the skill for general data analysis, reporting, statistics, and visualization rather than for a narrowly scoped CAN-bus or provenance purpose. Broad triggering increases the chance the skill is invoked on unrelated sensitive data and may expose users to unnecessary command execution, file writes, or network transmission.

Missing User Warnings

Medium

Confidence: 88% confidence
Finding: The documentation describes sending data to external OTS services, emitting webhook alerts, adapting HTTP and message-queue traffic, and exporting files, but does not prominently warn about privacy, data residency, network egress, or filesystem side effects. In an agent context, this can result in sensitive data being transmitted or persisted without informed consent or adequate safeguards.

Missing User Warnings

High

Confidence: 95% confidence
Finding: The skill advertises automatic rollback to the most recent verified snapshot without clearly warning that rollback is a destructive state-changing action that can overwrite legitimate updates or remove newer data. In automated environments, undocumented rollback behavior can cause integrity loss, service disruption, and difficult-to-recover data loss.

Static analysis

No suspicious patterns detected.