The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local prerequisite script .specify/scripts/bash/check-prerequisites.sh. This is a routine check within a structured project environment.
[COMMAND_EXECUTION]: The skill implements a 'hooks' system from .specify/extensions.yml where it may trigger other slash commands. The instructions specify that mandatory hooks must be explicitly executed and the AI must wait for results, while optional hooks are presented to the user for manual triggering. This is a standard extension pattern for project-specific automation.
[METADATA_POISONING]: The skill processes user-supplied $ARGUMENTS and data from .specify/memory/constitution.md and the specification file itself. However, it implements a highly structured clarification loop with a maximum of 5 questions, one-word/short-phrase constraints for answers, and explicit validation steps, which mitigates the risk of instructions embedded in data influencing the agent's core logic.

speckit-clarify