The Agent Skills Directory

[UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill requires the installation of the wxautox4 package and instructs the agent to clone an external code repository from https://github.com/cluic/wxauto-restful-api to function.
[DYNAMIC_EXECUTION]: The script scripts/wxapi.py uses subprocess.Popen to automatically launch the WeChat API service (run.py) from a dynamically determined directory if the service is not currently running.
[COMMAND_EXECUTION]: The skill executes shell commands via subprocess.Popen to start and manage the background API service, utilizing flags to detach the process on Windows and create new sessions on Linux.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill accesses local configuration files including ~/.wxautox/service_status.json and config.yaml to retrieve authentication tokens and connection parameters.
[INDIRECT_PROMPT_INJECTION]: The skill retrieves and displays WeChat messages from external contacts, creating an attack surface where malicious instructions embedded in messages could influence agent behavior.
Ingestion points: Message retrieval logic in cmd_getmsg, cmd_getmsg_chat, cmd_history, and cmd_newmsg in scripts/wxapi.py.
Boundary markers: Absent; message content is output to the agent context without delimiters or safety warnings.
Capability inventory: The skill possesses the capability to execute shell commands and launch subprocesses.
Sanitization: No sanitization or verification of the ingested message content is performed.

wxauto