The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes an external binary, zentao-cli, to perform operations on the ZenTao system. It executes shell commands such as zentao-cli execution list, create, get, update, and delete based on user requests, as documented across the references/ directory.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection (Category 8) because it passes user-supplied input directly to command-line arguments.
Ingestion points: Untrusted data enters the agent context through parameters like <name>, <project>, and <id> defined in files such as references/zentao-execution-create.md, references/zentao-execution-update.md, and references/zentao-execution-delete.md.
Boundary markers: No delimiters or specific instructions are provided to help the agent distinguish between command parameters and potential instructions embedded in user input.
Capability inventory: The zentao-cli tool has the capability to perform network operations and modify data within the ZenTao system across all functional reference files.
Sanitization: The skill does not define any sanitization, escaping, or validation logic for the input strings before they are utilized in the shell commands.

zentao-execution