The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands to interact with git and repository management tools (gh and glab). These commands detect the current repository configuration, summarize workspace changes, and manage context branches.
[INDIRECT_PROMPT_INJECTION]: The skill acts as a collector for untrusted data that is later processed by other AI agents in an autonomous pipeline.
Ingestion points: Untrusted data enters the context through user-provided issue descriptions and the contents of local workspace files read during the change detection process in references/workspace-changes.md.
Boundary markers: While Markdown headings and HTML <details> blocks are used to structure the issue body, there are no explicit delimiters or warnings to ignore potentially malicious instructions embedded in the gathered data.
Capability inventory: The skill possesses the ability to create GitHub issues (gh issue create) and push code changes to remote repositories (git push).
Sanitization: There is no evidence of sanitization, filtering, or escaping of the user-provided or file-based content before it is interpolated into the issue templates.

create-issue