Security audit

OPC Essentials

Security checks for vulnerabilities and agentic risk

Overview

This is a business productivity skill that gives structured drafts and operating guidance without executable code, persistence, credential use, or hidden data access.

Installers should expect this skill to shape business planning, email, invoice, lead, and client-management responses. Avoid pasting sensitive client, invoice, or relationship details unless you are comfortable having the assistant process that information in the current chat.

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

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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 (4)

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The skill advertises many broad, everyday situations like feeling overloaded, needing clarity, or drafting replies, which are common across general assistant use. That creates a genuine over-broad activation risk: the skill may trigger when the user did not specifically intend to use this business-ops workflow, causing context hijacking, irrelevant instruction injection, or unexpected shaping of responses.

Vague Triggers

Medium

Confidence: 92% confidence
Finding: The instruction to infer the best mode from context without clear boundaries gives the skill wide discretion to self-route based on ambiguous user input. In practice, this can cause unintended invocation or mode selection, especially when the user's request overlaps with generic planning, email drafting, or prioritization tasks outside the intended domain.

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The default routing logic maps very generic request patterns such as replying to someone, extracting next steps, choosing between options, or stepping back weekly to this skill. Because these patterns are common across many unrelated tasks, the router can over-capture user requests and steer outputs through this skill's business-oriented framing when that was not requested.

Vague Triggers

Medium

Confidence: 95% confidence
Finding: The example prompts are very broad, natural-language requests such as triage, prioritization, follow-up, and review assistance that closely resemble common user intents outside this specific skill. In an agent routing or skill-invocation system, this can cause the skill to activate too often, capture unrelated user requests, and expose business-sensitive workflow behavior or data handling in contexts where the user did not explicitly intend to use this premium operating-system skill.

Static analysis

No suspicious patterns detected.