Back to skill

Security audit

huawei-cloud-cbr-vault-list

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill suspicious with medium confidence: scanners report automatic collection of quality telemetry and installation of an unrelated CLI, which expand a read-only Huawei Cloud vault-listing task; the scan was partial.

Consider not installing until the telemetry behavior and the unrelated CLI installation are independently reviewed. The scan found 81.8% coverage, with two files partly inspected and some static checks degraded, so the evidence is incomplete.

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
  • Rogue AgentSelf-Modification, Session Persistence
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
Findings (14)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The skill makes an unrelated telemetry wrapper mandatory and directs automatic installation of executable software into the user's persistent PATH, expanding an inventory task into a software-installation trust decision. Although the dependency is disclosed in the body, installation lacks a separate approval boundary; a compromised release source could provide both a malicious package and its matching checksum.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

This duplicates the automatic dependency-installation concern: the read-only cloud inventory workflow introduces persistent executable installation and an additional release-service trust dependency. The documentation does mention the tool, so the issue is not complete concealment but mandatory scope expansion without a distinct user authorization step.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding

The commands concern managing skill-quality-cli rather than listing CBR vaults or configuring the declared hcloud/CBR SDK dependencies. This introduces an undeclared tool-management capability rather than an evident supporting detail of the stated purpose. The referenced installer scripts are not supplied, so their internal actions and any actual telemetry behavior cannot be established from this chunk.

Content

No source excerpt is available for this finding.

External Script Fetching

High
Category
Supply Chain
Confidence
90% confidence
Finding

Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.

Content

Scanner excerpt · scripts/install_cli.sh (reported line 28)May include surrounding context.

sh
API_URL="https://skillsapi.developer.myhuaweicloud.com/api/quality/cli/latest"

manifest="$(curl -fsSL --connect-timeout 5 --max-time 20 "${API_URL}")"
v="$(printf '%s' "$manifest" | python3 -c 'import sys,json;print(json.load(sys.stdin)["version"])')"
case "$(uname -s)" in
    Linux) ;;

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: env, shell.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

The skill adds automatic telemetry collection and dependency installation beyond the manifest's read-only vault-listing workflow.

Medium
Category
Excessive Agency
Confidence
96% confidence
Finding

Automatic telemetry and dependency installation introduce external processing and local changes beyond the requested vault inventory task. The opt-out variable does not provide affirmative consent, and the supplied material does not establish what the wrapper reports or whether sensitive command information is excluded.

Content

No source excerpt is available for this finding.

The guide introduces a telemetry-reporting wrapper that is unrelated to the declared vault-listing purpose.

Medium
Category
Excessive Agency
Confidence
93% confidence
Finding

The guide requires cloud commands to pass through an additional telemetry-reporting program without documenting its reported fields or obtaining opt-in. Such a wrapper runs in the credential-bearing environment and could observe inventory-related command information, although actual secret transmission is not demonstrated here.

Content

No source excerpt is available for this finding.

The skill automatically downloads and installs an unrelated quality CLI outside its declared vault-listing purpose.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The readiness helper automatically installs a mutable latest release of an unrelated executable into ~/.local/bin without a separate confirmation. SHA256 verification and archive checks mitigate corruption and path traversal, but the package and expected checksum both come from the same release manifest, leaving a compromised release source able to install attacker-controlled code for subsequent execution.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · scripts/ensure_cli.sh (reported line 89)May include surrounding context.

sh
echo "警告: 解压失败或 tar 包含不安全条目, 拒绝安装" >&2
        return 1
    fi
    mkdir -p ~/.local/bin/skill-quality-cli.d || { echo "警告: 创建安装目录失败" >&2; return 1; }
    [ -f "${tmp}/skill-quality-cli" ] || { echo "警告: 安装包内容缺失(skill-quality-cli)" >&2; return 1; }
    # 主二进制: 复制/可执行位失败必须显式失败(不再静默吞错)
    cp "${tmp}/skill-quality-cli" ~/.local/bin/skill-quality-cli || { echo "警告: 复制 skill-quality-cli 失败" >&2; return 1; }

The skill includes an unrelated quality-CLI downloader and persistent installer.

Medium
Category
Excessive Agency
Confidence
98% confidence
Finding

L26–L48 fetch release metadata and a package for skill-quality-cli, and L78–L88 install executable files and supporting Python modules under ~/.local/bin. This capability is unrelated to listing CBR vaults through the declared KooCLI or SDK paths and is not disclosed in the manifest. Although the script documents manual invocation and verifies the package checksum, those safeguards do not justify installing an unrelated tool as part of this skill.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · scripts/install_cli.sh (reported line 78)May include surrounding context.

sh
echo "错误: 解压失败或 tar 包含不安全条目, 拒绝安装" >&2
        exit 1
    fi
mkdir -p ~/.local/bin/skill-quality-cli.d
cp "${tmp}/skill-quality-cli" ~/.local/bin/skill-quality-cli || { echo "错误: 复制 skill-quality-cli 失败" >&2; exit 1; }
chmod +x ~/.local/bin/skill-quality-cli || { echo "错误: 设置可执行位失败" >&2; exit 1; }
if [ -f "${tmp}/skill-quality-cli.bin" ]; then

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/list_cbr_vaults.py (reported line 384)May include surrounding context.

python
cmd.extend(extra)
    cmd.extend(argv)
    try:
        proc = subprocess.run(
            cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE,
            timeout=120, universal_newlines=True,
        )

Tainted flow: 'cmd' from os.environ.get (line 378, credential/environment) → subprocess.run (code execution)

Medium
Category
Data Flow
Confidence
65% confidence
Finding

Data from a source is assigned to a variable that is later passed to a sink, creating a variable-mediated taint flow.

Content

Scanner excerpt · scripts/list_cbr_vaults.py (reported line 384)May include surrounding context.

python
cmd.extend(extra)
    cmd.extend(argv)
    try:
        proc = subprocess.run(
            cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE,
            timeout=120, universal_newlines=True,
        )

The script directly reads AK/SK credentials despite explicitly documenting that it never reads them itself.

Low
Category
Excessive Agency
Confidence
100% confidence
Finding

The module documentation at L007–L009 states that credentials are resolved by the hcloud profile or SDK environment and that this script never reads AK/SK itself. In contrast, L502–L519 directly retrieve access and secret keys from environment variables, and L593–L606 pass those values into BasicCredentials. This is a concrete contradiction about sensitive credential handling, although authentication itself is justified by the skill's purpose.

Content

No source excerpt is available for this finding.