The Agent Skills Directory

[COMMAND_EXECUTION]: Employs the system shell to execute gh CLI and git commands for repository introspection, label creation, and issue management.
[COMMAND_EXECUTION]: Uses mktemp and rm for secure handling of temporary files during the issue generation process.
[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by reading untrusted data from the repository (e.g., existing issue titles, commit messages, and configuration files) to inform its architectural analysis.
Ingestion points: SKILL.md reads file contents (e.g., package.json, README.md), existing issue metadata, and commit logs.
Boundary markers: No explicit delimiters are used when interpolating repository content into analysis prompts.
Capability inventory: Shell execution capability via the gh and git command-line tools.
Sanitization: No explicit sanitization or escaping of the ingested data is described before it is used to generate the issue title or body content.

generate-gh-issue