The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or safety bypass attempts were detected. The skill follows security best practices by enforcing explicit human approval for all state-changing operations and disabling autonomous model-initiated tool calls.
[COMMAND_EXECUTION]: Employs gh (GitHub CLI) and git for repository management. All network mutations (e.g., creating/deleting labels, patching repository settings) require the user to explicitly approve a generated plan before any command is executed.
[DATA_EXFILTRATION]: Network activity is confined to the official GitHub API for legitimate administrative purposes. The skill does not access sensitive credential files (like SSH keys or AWS configs) or transmit data to unauthorized external domains.
[PROMPT_INJECTION]: The skill reads from local instruction files (CLAUDE.md, AGENTS.md) and GitHub issue descriptions, which constitutes an indirect prompt injection surface. This risk is effectively mitigated by structural boundaries (Markdown headers), mandatory manual review of the proposed execution plan, and the disable-model-invocation: true setting.
[REMOTE_CODE_EXECUTION]: The included Python script (labels.test.py) is a static validation utility for checking documentation compliance. It does not perform remote code fetching, dynamic execution, or unsafe deserialization.

setup-dividedby-skills