The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands (curl) to interact with the Feishu API. It also uses inline Python scripts to safely calculate message signatures and format JSON payloads, which is a recommended practice to avoid shell injection vulnerabilities.
[DATA_EXFILTRATION]: The skill transmits data to Feishu's official domains (open.feishu.cn). While it handles message content and files, this behavior is entirely consistent with its documented purpose as a messaging tool and targets a well-known enterprise service.
[CREDENTIALS_UNSAFE]: The skill requires several sensitive credentials (API keys and secrets) but correctly instructs the user to provide them via environment variables rather than hardcoding them into the instructions or scripts.

send-feishu