The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes standard GitHub CLI (gh) commands to manage workflow specifications. It performs environment checks (gh auth status), repository metadata retrieval (gh repo view), and issue management (gh issue create). These commands are used for their intended purpose within the current project scope.
[DATA_EXPOSURE]: The skill reads a local file named NOTES.md at the project root. This file is used to store context about the user's tools and terminology. While the skill processes this data, it is used locally to inform the interview process and populate the GitHub issues as requested by the user.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from a local file (NOTES.md).
Ingestion points: Reads NOTES.md in Phase 1 to identify recurring processes.
Boundary markers: None explicitly defined in the instructions for file reading.
Capability inventory: The skill has the capability to execute shell commands (gh) and write to external repositories (GitHub Issues).
Sanitization: The skill involves a human-in-the-loop 'grilling' process where the user must confirm options and details before any issue is published, significantly mitigating the risk of processing malicious content from the notes file.

loop-me