ClawHub
Back to skill

Security audit

Aws Toolkit

Security checks across malware telemetry and agentic risk

Overview

This AWS operations skill is mostly coherent, but it normalizes high-impact cloud changes without enough confirmation, scoping, or credential-safety guidance.

Install only if you are comfortable with an agent helping run AWS administration commands. Use a non-production account or tightly scoped IAM role first, avoid long-lived keys, confirm account and region before any apply/deploy action, and remove or override any --auto-approve use unless you explicitly intend unattended changes.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (7)

Vague Triggers

Medium
Confidence
87% confidence
Finding
The skill description uses very broad activation criteria such as security detection, compliance audit, vulnerability scanning, and encryption protection, which can cause an agent to invoke a powerful AWS management skill in situations where the user did not clearly request infrastructure or cloud operations. Because the skill also exposes exec-capable workflows, ambiguous triggering increases the chance of unintended high-impact actions in a sensitive cloud environment.

Missing User Warnings

High
Confidence
96% confidence
Finding
The example shows infrastructure deployment with an apply step using --auto-approve, but does not provide a prominent warning that it can create or modify paid, production-impacting AWS resources across multiple regions. In an agent context, examples often become operational defaults, so this normalizes destructive or costly behavior without requiring explicit user acknowledgment.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill documents required AWS credentials and profiles but does not include a strong privacy and security warning about handling secrets, avoiding logging them, and limiting their scope. In an agent environment, this omission can lead to unsafe prompting, credential exposure in output, or use of overly privileged keys during automated cloud operations.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
# 部署基础设施
python3 scripts/infra.py apply \
  --config ./terraform/ \
  --auto-approve

# 输出：
# VPC: 10.0.0.0/16 (2区域)
Confidence
98% confidence
Finding
auto-approve

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
### 命令参数说明

- `--apply-recommendations`: 命令参数,用于指定操作选项
- `--auto-approve`: 命令参数,用于指定操作选项
- `--period`: 命令参数,用于指定操作选项
- `--dry-run`: 命令参数,用于指定操作选项
- `-DSS`: 命令参数,用于指定操作选项
Confidence
89% confidence
Finding
auto-approve

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
# 部署基础设施
python3 scripts/infra.py apply \
  --config ./terraform/ \
  --auto-approve

# 输出：
# VPC: 10.0.0.0/16 (2区域)
Confidence
98% confidence
Finding
--auto-approve

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
### 命令参数说明

- `--apply-recommendations`: 命令参数,用于指定操作选项
- `--auto-approve`: 命令参数,用于指定操作选项
- `--period`: 命令参数,用于指定操作选项
- `--dry-run`: 命令参数,用于指定操作选项
- `-DSS`: 命令参数,用于指定操作选项
Confidence
89% confidence
Finding
--auto-approve

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.