awesome-openclaw-usecases-zh

Warn

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: MEDIUMDATA_EXFILTRATIONCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill provides Python code snippets for accessing and reading the local WeChat messaging database file (WeChat/Msg/Multi/MSG0.db) to extract chat history for persona distillation. Accessing local messaging databases is a high-risk activity that can lead to significant data exposure if executed without strict oversight.\n- [CREDENTIALS_UNSAFE]: The documentation describes a method to programmatically extract authentication cookies from a user's browser using the browser-cookie3 library to maintain session authentication. This technique involves harvesting sensitive credentials from the local filesystem.\n- [COMMAND_EXECUTION]: The skill demonstrates the use of command-line tools, such as the lark-cli, to interact with the Feishu platform for document searching and messaging operations.\n- [EXTERNAL_DOWNLOADS]: The skill provides instructions to fetch reference material from public repositories on GitHub and AtomGit (AlexAnys/awesome-openclaw-usecases-zh).
Audit Metadata
Risk Level
MEDIUM
Analyzed
May 16, 2026, 08:18 PM
Security Audit — agent-trust-hub — awesome-openclaw-usecases-zh