100Critical
This skill is highly vulnerable to command injection and prompt injection
Behavior Analysis
Description verified — behavior matches claim
Claims to do
腾讯云 COS 技能: 一站式管理腾讯云对象存储(COS)和数据万象(CI)，通过统一的 Node.js SDK 脚本提供以下能力：
Actually does
This skill executes Node.js and bash scripts to interact with Tencent Cloud COS (Object Storage) and CI (Data Vison) services. It performs extensive operations including file uploads/downloads, bucket management, image/document/media processing, content moderation, speech recognition/synthesis, file utilities, and MetaInsight search capabilities. It manages Tencent Cloud API credentials via environment variables or encrypted local files, contacting Tencent Cloud APIs for all operations.
Skill Info
Nameshawnminh/tencent-cloud-cos
Registryclawhub
Version1.1.1
PURLpkg:clawhub/shawnminh/tencent-cloud-cos@1.1.1
Downloads1,944
Installs2
Source
SHA-256e3b5ef320311...
SHA-1b2547777309f...
MD54830a04aeb0d...
TLSH60c208e99cc2...
ssdeep768:6mutHQ73...
Size27,496 bytes
Install
This skill has been flagged as potentially malicious. Review the findings below before installing.
ClawHubDocs
openclaw skills install shawnminh/tencent-cloud-cosAssessments (2)
AI Agent Traps ↗Prompt Injection Detectedcriticalprompt injection
DeBERTa classifier detected prompt injection (confidence: 0.98)
98% confidence
Generic Cloud API Call Capabilitycriticaltool poisoning
The 'ci-request' action allows making arbitrary calls to the Tencent Cloud CI API by specifying method, path, and body. This provides a broad attack surface, enabling actions not explicitly defined or potentially bypassing intended restrictions.
90% confidenceline 573
ci-request --method POST --path "image/auditing" --body '<xml>...</xml>'
Potential Command Injection via Script Argumentshighcommand execution
The skill executes shell scripts (`setup.sh`) and Node.js scripts (`cos_node.mjs`) with user-provided arguments (e.g., file paths, keys, content, bucket names). Without robust input sanitization in the underlying scripts, this presents a significant risk for command injection.
80% confidenceline 109
{baseDir}/scripts/setup.sh --check-only
node {baseDir}/scripts/cos_node.mjs <action> [--option value ...]Credential Persistence on Diskhighpersistence
The skill offers a '--persist' option to write cloud credentials to a local '.env' file and an 'encrypt-env' action to encrypt them to '.env.enc'. While security measures like file permissions and encryption are mentioned, any disk persistence of credentials increases the attack surface.
90% confidenceline 130
setup.sh --from-env --persist node scripts/cos_node.mjs encrypt-env
Automated Cloud Resource Creation (Resource Abuse)mediumresource abuse
The 'create-knowledge-base' action automatically creates cloud resources (storage bucket, dataset). If an attacker can repeatedly trigger this action, it could lead to resource exhaustion or unexpected cloud billing charges.
70% confidenceline 399
create-knowledge-base --name <用户指定的名称> 自动完成三步：创建存储桶 → 创建 DocSearch 数据集 → 绑定
Badge
Markdown
[](https://mondoo.com/ai-agent-security/skills/clawhub/shawnminh/tencent-cloud-cos)HTML
<a href="https://mondoo.com/ai-agent-security/skills/clawhub/shawnminh/tencent-cloud-cos"><img src="https://mondoo.com/ai-agent-security/api/badge/clawhub/shawnminh/tencent-cloud-cos.svg" alt="Mondoo Skill Check" /></a>Image URL
https://mondoo.com/ai-agent-security/api/badge/clawhub/shawnminh/tencent-cloud-cos.svgSecure your AI agents
Skills can read files, run commands, and access credentials. Mondoo helps organizations manage the security risks of AI agent skills across their entire fleet.
- Continuous scanning of skills across all registries
- Policy enforcement before skills reach your agents
- Integration with your existing security workflow