The Agent Skills Directory

[COMMAND_EXECUTION]: Utilizes standard Git and GitHub CLI (gh) tools via Bash to perform repository operations, which is the primary intended behavior of the skill.
[PROMPT_INJECTION]: Vulnerable to indirect prompt injection via the .chalk/chalk.json file. Malicious commands or instructions placed in the test.command or dev.command fields could be incorporated into the generated PR description.
Ingestion points: Reads project-specific metadata from .chalk/chalk.json at runtime.
Boundary markers: Absent. The instructions do not specify any delimiters or safety prompts to prevent the agent from following instructions embedded in the external data.
Capability inventory: Uses Bash for command execution and repository synchronization via git push and gh pr create.
Sanitization: Uses HEREDOC to pass data to the shell, providing protection against command injection but not against semantic instruction injection within the PR body.

create-pr