everything-claude-code-zh

Warn

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to clone an external repository from https://github.com/xu-xiang/everything-claude-code-zh.git to begin installation.
  • [REMOTE_CODE_EXECUTION]: Users are directed to run ./install.sh immediately after cloning the repository, which executes unverified shell scripts directly on the host machine.
  • [COMMAND_EXECUTION]: The skill performs several potentially dangerous system operations:
  • Modifies file permissions using chmod +x on downloaded JavaScript hooks.
  • Installs global system packages using npm install -g pm2.
  • Performs recursive directory copies into sensitive user configuration paths like ~/.claude/rules/ and ~/.claude/agents/.
  • [DATA_EXFILTRATION]: The configuration section encourages the export of AGENTSHIELD_API_KEY, which is used for a third-party security scanning service that may transmit project data externally.
  • [PROMPT_INJECTION]: The skill exhibits a significant indirect prompt injection surface as it ingests untrusted data via the market-research and article-writing skills while possessing high-privilege capabilities such as file system writes and script execution in the background.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 17, 2026, 06:51 PM
Security Audit — agent-trust-hub — everything-claude-code-zh