Security audit

Sui Coverage

Security checks across malware telemetry and agentic risk

Overview

This is a mostly coherent Sui Move coverage helper, but it tells agents to always stage and commit repository changes without an explicit approval step.

Install only if you are comfortable with an agent running local Sui coverage commands and proposing or editing tests. Before use, instruct the agent not to stage or commit anything without explicit approval, and review diffs before accepting generated tests or audit findings.

SkillSpector

By NVIDIA

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration

Findings (5)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 89% confidence
Finding: The skill advertises and documents use of shell commands, file reads, and file writes, but the metadata declares no permissions or safety boundaries. This creates a transparency and consent problem: an agent or user may invoke the skill expecting passive analysis while it can modify repository contents and execute external tooling.

Tp4

High

Category: MCP Tool Poisoning
Confidence: 95% confidence
Finding: The documented behavior overstates the skill's capabilities by claiming it can write missing tests and perform security audits, while the described tooling primarily runs coverage commands and generates reports. This mismatch is dangerous because users may trust the skill to make code changes or provide security assurance it cannot actually deliver, leading to unsafe reliance and unreviewed repository modifications through adjacent workflow instructions.

Context-Inappropriate Capability

Low

Confidence: 82% confidence
Finding: The skill instructs agents to run git add and git commit, extending behavior from analysis into repository modification and version-control operations. Even if common in development workflows, committing automatically can persist unwanted or incorrect changes and make rollback harder, especially when done without explicit user approval.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The workflow tells the agent to write missing tests automatically but does not prominently warn that repository files will be modified. In an agent setting, silent file modification is risky because generated tests may be incorrect, may encode wrong assumptions, or may alter the project in ways the user did not authorize.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The instruction to always commit test improvements normalizes irreversible version-control changes without warning or consent. In practice this can turn an exploratory analysis skill into one that persists generated content and history changes automatically, increasing the blast radius of mistakes or prompt-induced bad edits.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.