The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the gh (GitHub) CLI tool to create issues. It constructs the command using a 'slug' derived from user input. The instruction to 'Generate a kebab-case slug' serves as a sanitization step to prevent shell command injection, and the operation is gated by an interactive user prompt.
[DATA_EXFILTRATION]: The skill can transmit the generated PRD content to GitHub's servers. This is an intended feature that utilizes a well-known service and requires explicit user consent before proceeding.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests and processes untrusted data from both user input and local project files to generate its output.
Ingestion points: Project documentation (README.md, CLAUDE.md, architecture docs) and the user-provided product idea.
Boundary markers: None identified in the prompt instructions.
Capability inventory: Local file reads and shell command execution via the gh utility.
Sanitization: Employs kebab-case string transformation for arguments used in shell commands.

shape