The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes structured data editors such as jq, yq, tomlq, and xq to perform precise, format-aware edits on JSON, YAML, TOML, and XML files.
[COMMAND_EXECUTION]: The instructions direct the agent to execute repository-native verification tools, including formatters, linters, and test suites, to ensure change correctness after each patch.
[SAFE]: The skill implements a disciplined 'Inspect-Plan-Patch-Verify' loop designed to minimize the risk of accidental deletions, speculative rewrites, or architectural drift.
[SAFE]: Specific guardrails are provided to protect sensitive areas, including credential-bearing files (.env, .aws), infrastructure definitions, and database migrations, by requiring minimal changes and explicit review.

edit-files