The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions for the agent to clone the API service repository from https://github.com/cluic/wxauto-restful-api.git if the service is not found locally. This is a vendor-owned resource used to provide the skill's core functionality.
[COMMAND_EXECUTION]: The scripts/wxapi.py script uses subprocess.Popen to launch the API service (run.py) on the local machine. Additionally, the skill instructs the agent to execute shell commands for environment setup, including git clone, pip install, and creating Python virtual environments.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data in the form of WeChat chat messages via the getmsg and newmsg commands. This creates a vulnerability where malicious instructions sent by WeChat users could be interpreted by the agent as commands.
Ingestion points: Data is retrieved in scripts/wxapi.py via HTTP requests to the WeChat API service.
Boundary markers: No explicit delimiters or instructions are used in the script to separate untrusted message content from system prompts.
Capability inventory: The skill possesses the ability to execute shell commands, perform network requests, and read local configuration files.
Sanitization: No evidence of sanitization or filtering of the ingested message content was found.

wxauto