Security audit

multi task

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed parallel task orchestration guide, but users should be aware it can spawn multiple subagents and retry failed work automatically.

Install this only if you want the agent to coordinate parallel subagents for batch work. Before dispatch, review the listed work units, input paths, output paths, and whether retries or background execution are acceptable for the task.

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

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Vague Triggers

High

Confidence: 96% confidence
Finding: The skill description is triggered by extremely common language such as 'each', 'every', 'all', numbered lists, folders, and any decomposable 3+ item workflow, which can cause the agent to invoke this skill in many situations far beyond clearly intended batch orchestration. Because the skill then encourages creating and dispatching multiple subagents and reading shared context, overbroad activation can amplify actions, increase data exposure across prompts, and lead to unintended large-scale operations without sufficiently specific user intent.

Natural-Language Policy Violations

Medium

Confidence: 90% confidence
Finding: The skill instructs the agent to retry failed tasks automatically up to two times, which authorizes repeated system actions without renewed user approval. In a skill designed for parallel execution, this can multiply unintended writes, repeated external operations, or repeated access to sensitive files across many tasks, reducing user control and making errors costlier.

Static analysis

No suspicious patterns detected.