Security audit

IRIS代码格式化免费版

Security checks for vulnerabilities and agentic risk

Overview

This is a Markdown-only IRIS/ObjectScript formatter skill, but it asks for command and write capability while describing broad development/deployment triggers and create/modify/delete/save operations beyond simple formatting.

Install only if you intend to use it for IRIS/ObjectScript formatting or style review. Keep use scoped to files you choose, review any edits before applying them, and avoid letting it run broad shell commands or deployment-related tasks unless you explicitly requested and understand them.

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Behavioral ASTexec() Call, eval() Call, Dynamic Import
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

Findings (5)

Description-Behavior Mismatch

Medium

Confidence: 91% confidence
Finding: The skill is presented as a formatter and standards checker, but the documented capabilities expand into create/modify/delete, save, export, and conversion actions. This capability drift increases the chance that an agent or user invokes the skill for state-changing operations that were not clearly scoped or safety-reviewed, especially since the manifest also allows execution tools.

Description-Behavior Mismatch

Medium

Confidence: 84% confidence
Finding: The metadata says the skill is not suitable for operations/deployment management, yet other text says to use it during development/deployment scenarios. Contradictory activation guidance can cause the agent to invoke the skill in higher-risk contexts than intended, including environments with credentials, servers, or production-adjacent code.

Vague Triggers

High

Confidence: 93% confidence
Finding: The top-level description says to use the skill for broad categories like code generation, programming assistance, debugging, testing, and development deployment. Such expansive trigger language makes accidental invocation likely across many developer workflows, increasing the chance the skill gains access to sensitive code, executes commands, or influences tasks outside simple formatting.

Vague Triggers

High

Confidence: 92% confidence
Finding: The dedicated trigger section is vague and broad, stating the skill should be used during code generation, programming assistance, debugging, testing, and development deployment. In a tool-enabled agent, ambiguous triggers can route unrelated requests into a skill with exec capability, raising the risk of unnecessary command execution or unauthorized code modification.

Missing User Warnings

Medium

Confidence: 89% confidence
Finding: The skill describes modify/delete/save/export-style operations but does not provide clear user-facing warnings that these actions may alter code or files. In combination with write and exec permissions, this can lead to silent or unexpected state changes, especially if an agent interprets these operations as approved by default.

Static analysis

No suspicious patterns detected.