The Agent Skills Directory

[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it processes untrusted data from code diffs, PR summaries, and documentation files to generate its questions.
Ingestion points: Data is ingested via git diff output, gh CLI tool output, and local file reads from the docs/superpowers/ directory.
Boundary markers: The instructions do not define delimiters or provide warnings to the agent to ignore instructions contained within the artifact data.
Capability inventory: The agent can execute shell commands (git, gh) and read files in the repository.
Sanitization: The skill does not perform any sanitization or filtering of the artifact content to prevent embedded instructions from overriding the agent's behavior.
[COMMAND_EXECUTION]: The skill instructs the agent to execute git and gh (GitHub CLI) commands. These tools are used for their intended purpose of reading repository state and are well-known developer utilities.

quiz-me