The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a local binary wecom-cli to perform tasks. Inputs are passed via a structured JSON parameter, which is a secure method of command execution.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text for task details. The vulnerability surface is mitigated by high-priority security instructions that forbid the agent from treating tool output as instructions.
[SAFE]: The skill implements safety measures such as mandatory pre-checks from a shared skill and instructions to protect internal system tokens and IDs from user exposure.

wecomcli-todo