The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local setup script located at .specify/scripts/bash/setup-plan.sh to extract project parameters.
[DYNAMIC_EXECUTION]: The skill reads the .specify/extensions.yml file and dynamically executes arbitrary shell commands defined in the hooks.before_plan and hooks.after_plan sections via the EXECUTE_COMMAND directive.
[INDIRECT_PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted data and uses it to drive shell execution.
Ingestion points: The skill reads input from $ARGUMENTS, the .specify/extensions.yml configuration file, the JSON output of the setup script, and the project's feature specification files (SKILL.md).
Boundary markers: Absent. There are no explicit delimiters or warnings to ignore instructions embedded within the data read from these external sources.
Capability inventory: The skill possesses the capability to execute local bash scripts and arbitrary shell commands through its hook system (SKILL.md).
Sanitization: Absent. There is no evidence of input validation or sanitization for the data ingested from user arguments or local project files before it influences command execution.

speckit-plan