The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the GitHub CLI (gh) to manage issues and interact with the GitHub REST API. Specifically, it uses:
- gh issue view to retrieve task descriptions from parent issues.
- gh issue create and gh issue edit to build and structure implementation plans.
- gh api to programmatically link child issues as native sub-issues of the parent. These commands are localized to the repository of the current working directory and align with the skill's stated purpose.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface by processing data from existing GitHub issues.
- Ingestion points: The task description is retrieved from a parent GitHub issue via the gh issue view command as described in SKILL.md.
- Boundary markers: The skill appends metadata to the issue body using a --- horizontal rule as a separator, but it does not implement explicit delimiters or instructions for the model to ignore embedded commands within the ingested task text.
- Capability inventory: The agent has the capability to create new repository content (gh issue create) and modify existing issues (gh issue edit).
- Sanitization: There is no evidence of automated filtering or sanitization of the issue content before it is used to generate new issues. This risk is addressed by a mandatory human-in-the-loop requirement; the skill explicitly stops for user review and approval of the proposed plan before any issues are created or modified.

codely-plan-create-github