The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes standard system utilities including git and gh (GitHub CLI) for repository history analysis and issue management. Notable implementations include the use of git bisect run to automate regression testing and git log with explicit pattern-type flags to ensure consistent behavior across different environment configurations.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input in the form of bug descriptions and reproduction details. This information is used to guide codebase exploration, generate reproduction scripts, and populate GitHub issues. This represents a potential injection surface; however, the risk is mitigated by design patterns that require the agent to present hypotheses and plan summaries to the user for verification before execution or persistence.
[SAFE]: The skill's external references point to well-known academic and professional literature in the fields of systems thinking and software engineering (e.g., works by Donella Meadows and Andreas Zeller). The instructions prioritize deterministic reproduction and systematic diagnosis over speculative code changes.

triage-issue