Back to skill

Security audit

huawei-cloud-eip-cost-optimizer

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly reports on Huawei Cloud EIPs, but its documentation and helper scripts include host-modifying setup, cron persistence, and write-capable IAM/API guidance that exceed the read-only framing.

Install only if you are comfortable with an EIP monitoring tool that may install local dependencies, create audit logs, and optionally add a cron job. Use the minimum read-only IAM policy only; do not grant the full-management policy for this skill. Avoid the curl-to-bash installer path unless you independently verify the installer source, and review any --fix or --setup-cron command before running it.

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
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (64)

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

The documentation claims a read-only cloud analysis skill, yet it includes dependency installation, environment remediation, and auto-fix flows that modify the local system. This misrepresentation can cause operators to authorize execution in sensitive environments where host modification is not acceptable.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The documentation claims a read-only cloud analysis skill, yet it includes dependency installation, environment remediation, and auto-fix flows that modify the local system. This misrepresentation can cause operators to authorize execution in sensitive environments where host modification is not acceptable.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

The documentation claims a read-only cloud analysis skill, yet it includes dependency installation, environment remediation, and auto-fix flows that modify the local system. This misrepresentation can cause operators to authorize execution in sensitive environments where host modification is not acceptable.

Content

No source excerpt is available for this finding.

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

Piping a remote script directly from curl into bash executes unverified code from the network with no integrity check, review, or pinning. In an agent or automation environment, that creates a strong supply-chain compromise path where a tampered server, MITM, or replaced script leads to arbitrary code execution.

Content

Scanner excerpt · SKILL.md (reported line 54)May include surrounding context.

bash
# Linux/macOS one-click install
curl -sSL https://hwcloudcli.obs.cn-north-1.myhuaweicloud.com/cli/latest/hcloud_install.sh | bash

# Verify installation
hcloud --version

Chaining Abuse

High
Category
Tool Misuse
Confidence
98% confidence
Finding

The command chains network retrieval directly into shell execution, eliminating any opportunity to inspect content before it runs. This pattern is especially dangerous in skills because it can be reused as a one-step arbitrary code execution primitive under the guise of dependency setup.

Content

Scanner excerpt · SKILL.md (reported line 54)May include surrounding context.

bash
# Linux/macOS one-click install
curl -sSL https://hwcloudcli.obs.cn-north-1.myhuaweicloud.com/cli/latest/hcloud_install.sh | bash

# Verify installation
hcloud --version

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 76)May include surrounding context.

md
- `scripts/list_eips.sh` - List all EIPs in a region (supports filtering and summary)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 130)May include surrounding context.

md
- `scripts/list_eips.sh` - List all EIPs in a region (supports filtering and summary)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 183)May include surrounding context.

md
- `scripts/list_eips.sh` - List all EIPs in a region (supports filtering and summary)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 195)May include surrounding context.

md
- `scripts/list_eips.sh` - List all EIPs in a region (supports filtering and summary)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 198)May include surrounding context.

md
- `scripts/list_eips.sh` - List all EIPs in a region (supports filtering and summary)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 201)May include surrounding context.

md
- `scripts/list_eips.sh` - List all EIPs in a region (supports filtering and summary)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 336)May include surrounding context.

md
- `scripts/list_eips.sh` - List all EIPs in a region (supports filtering and summary)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 337)May include surrounding context.

md
- `scripts/list_eips.sh` - List all EIPs in a region (supports filtering and summary)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 409)May include surrounding context.

md
- `scripts/list_eips.sh` - List all EIPs in a region (supports filtering and summary)

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
85% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · references/cli-installation-guide.md (reported line 34)May include surrounding context.

bash
curl -sSL -o /tmp/hcloud_install.sh https://hwcloudcli.obs.cn-north-4.myhuaweicloud.com/cli/latest/hcloud_install.sh
bash /tmp/hcloud_install.sh
rm -f /tmp/hcloud_install.sh

Method 3: Package Manager

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
85% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · references/cli-installation-guide.md (reported line 123)May include surrounding context.

bash
curl -sSL -o /tmp/hcloud_install.sh https://hwcloudcli.obs.cn-north-4.myhuaweicloud.com/cli/latest/hcloud_install.sh
bash /tmp/hcloud_install.sh
rm -f /tmp/hcloud_install.sh

Method 3: Package Manager

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
85% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · references/cli-installation-guide.md (reported line 132)May include surrounding context.

Uninstall

bash
rm -f /usr/local/bin/hcloud
rm -rf ~/.hcloud

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
90% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · references/cli-installation-guide.md (reported line 133)May include surrounding context.

bash
rm -f /usr/local/bin/hcloud
rm -rf ~/.hcloud

Related Documentation

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
90% confidence
Finding

Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).

Content

Scanner excerpt · references/cli-installation-guide.md (reported line 133)May include surrounding context.

bash
rm -f /usr/local/bin/hcloud
rm -rf ~/.hcloud

Related Documentation

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill metadata explicitly states a read-only analysis-only scope, but this document declares and distributes a full-management permission set including create, delete, update, associate, and disassociate actions. This creates a clear scope mismatch that can enable destructive or costly changes if the skill, a downstream agent, or a compromised operator follows the broader policy instead of the stated intent.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The listed capabilities allow creation, deletion, modification, and rebinding of EIPs, none of which are justified by a cost-reporting and audit skill advertised as non-mutating. If granted, these permissions could be abused to release public IPs, alter network exposure, or incur additional cloud costs, turning a read-only helper into an infrastructure-changing actor.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill is explicitly described as read-only analysis only, but this code can download, chmod, and execute an installation script when --fix is used. Even with a minimal grep-based check, executing remotely fetched code introduces a strong supply-chain and arbitrary code execution risk that exceeds the declared scope.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The code downloads a shell script from a remote URL and executes it locally. The presence of only a non-empty-file check and a grep for the string 'hcloud' is not meaningful integrity verification, so a compromised host, mirror, or network path could deliver malicious code for execution.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The script invokes apt-get/yum to install jq and bc, which modifies the local system and violates the skill's stated read-only boundary. This expands the skill from inspection into system mutation, creating unnecessary privilege and integrity risk if run in sensitive environments.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The script implements local crontab installation even though the skill is described as read-only analysis only. This breaks the stated trust boundary by persisting execution on the host, creating unauthorized scheduler modification and ongoing execution capability beyond simple EIP inspection.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.