The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from an external social platform (topics, questions, and comments) via zsxq-topic-detail.md and zsxq-topic-search.md. This content could contain instructions designed to manipulate the agent. While the skill includes instructions to seek user confirmation before performing actions, it lacks explicit boundary markers (like XML tags or delimiters) to differentiate between user data and system instructions.
[COMMAND_EXECUTION]: The skill relies on the execution of shell commands via the zsxq-cli binary to perform its functions. These commands are used for searching, viewing, creating, editing, and deleting topics. The skill documentation explicitly requires user confirmation for destructive or public-facing commands, which mitigates the risk of unauthorized execution.

zsxq-topic