The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The skill recommends using chmod +x to grant execution permissions to hook scripts, which may include scripts from external or third-party plugins.\n- [COMMAND_EXECUTION]: The skill provides detailed instructions on defining and troubleshooting shell commands and processes that the agent is expected to execute via the hook system.\n- [DYNAMIC_EXECUTION]: Describes a mechanism for runtime execution of commands and scripts triggered by platform events, using either shell strings or direct process calls.\n- [INDIRECT_PROMPT_INJECTION]: The skill guides the agent to read and act upon hook configuration data which could originate from untrusted sources such as repository plugins.\n
Ingestion points: Reads hook definitions from ~/.zcode/cli/config.json and plugin hooks/hooks.json files.\n
Boundary markers: No explicit delimiters or safety warnings are used when processing the data from these configuration sources.\n
Capability inventory: The skill utilizes the ability to execute shell commands, spawn processes, and modify file system permissions.\n
Sanitization: No sanitization or validation of the commands defined in the configuration files is described.

diagnosing-hooks