The Agent Skills Directory

[COMMAND_EXECUTION]: The skill leverages the weread-omni CLI for all operations, ensuring structured interaction with WeRead data through JSON-formatted shell commands.
[DATA_EXFILTRATION]: Security instructions explicitly prohibit the display of raw authentication tokens or credentials, and limit article retrieval to a validated WeChat domain with no authentication headers sent to that host.
[INDIRECT_PROMPT_INJECTION]: The skill mitigates risks associated with processing external articles and book text by requiring explicit user confirmation and specific command-line flags for any destructive actions like deletions.

weread