本地文件上传云端

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:30 PM UTC
• Version:
  1.0.0

Overview

This skill does what it says by uploading a user-selected local file to Qiniu Cloud Storage, but users should treat uploaded files as potentially public.

Install only if you intend to upload selected local files to Qiniu Cloud Storage. Do not use it for private or confidential files, verify the Qiniu bucket/domain and credential scope first, and install the qiniu package from a trusted source.

SkillSpector

By NVIDIA

Vulnerability Patterns

• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (2)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 94% confidence
• Finding: The skill performs sensitive capabilities—reading local files and making network uploads—without declaring permissions, which undermines transparency and informed consent for users or calling systems. In this context, the risk is real because the skill uploads arbitrary local files to a cloud bucket and can expose their contents externally, so missing permission declarations materially increase the chance of unintended data disclosure.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 91% confidence
• Finding: The stated behavior says the skill uploads a file and returns a URL, but the implementation also reads ~/.openclaw/config.json to obtain credentials and storage settings, which is a security-relevant side effect not disclosed in the description. This mismatch is dangerous because hidden access to local configuration/secrets reduces user awareness and could expose cloud credentials or other sensitive configuration data if the skill is modified, misused, or broadly trusted based on incomplete documentation.

VirusTotal

67/67 vendors flagged this skill as clean.