The Agent Skills Directory

[COMMAND_EXECUTION]: Uses Bash shell scripts with "curl" and "jq" to interact with the official WeCom API. The script structure is secure and relies on robust JSON parsing rather than unsafe string manipulation.
[CREDENTIALS_UNSAFE]: Accesses the WECOM_CORP_SECRET environment variable for authentication. The skill includes explicit instructions to the agent to treat this secret as a password and never leak it in logs or user-facing responses.
[DATA_EXFILTRATION]: Performs network operations solely to the official WeCom API endpoint (qyapi.weixin.qq.com). No suspicious or unauthorized third-party connections were identified.
[PROMPT_INJECTION]: Contains proactive safety instructions that prevent the agent from sending messages or creating meetings without explicit user approval, mitigating risks associated with accidental or malicious instruction obedience.

wecom