The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it reads existing content from GitHub issues to resume brainstorming sessions.
Ingestion points: The skill uses gh issue view (in SKILL.md and issue-lifecycle.md) to retrieve the body and log of existing issues.
Boundary markers: There are no explicit delimiters or instructions to the agent to ignore or treat the retrieved issue body as untrusted data.
Capability inventory: The skill has the capability to execute gh issue create and gh issue edit, which can modify the repository's issue tracker.
Sanitization: The skill does not perform any escaping, validation, or filtering of the content retrieved from GitHub before processing it.
[COMMAND_EXECUTION]: The skill relies on the GitHub CLI (gh) for its core functionality. While it recommends using --body-file for large updates, some recipes in issue-lifecycle.md demonstrate string interpolation for issue titles (e.g., gh issue edit 47 --title "[DRAFT] <type(scope): summary>"), which could be a surface for command injection if the agent does not properly escape user-provided values when constructing the shell command.

brainstorming-to-issue