The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses several shell-based tools including gh (GitHub CLI), git, grep, ls, and find to gather repository context and perform the final issue creation. This is standard behavior for developer-oriented automation.
[DATA_EXFILTRATION]: Local repository data (source code snippets, logs, and issue history) is read to construct issue reports. These reports are then transmitted to GitHub's servers via the gh issue create command. This behavior is documented as the skill's primary purpose and is gated by a mandatory user confirmation step.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests untrusted data from existing GitHub issues and the local codebase to "infer conventions."
Ingestion points: Reads existing issue titles and bodies from GitHub (SKILL.md, Steps 2 and 5) and repository history via git log (SKILL.md, Step 4).
Boundary markers: None identified; the agent is instructed to follow the conventions found in the ingested data.
Capability inventory: Includes file system exploration (ls, find), content searching (grep), and GitHub issue creation (gh issue create).
Sanitization: Employs shell heredocs (EOF) during the final publication step (SKILL.md, Step 9) to prevent command injection when passing the generated body text to the shell.

github-create-issue