Use when creating, editing, validating, benchmarking, or packaging agent skills, SKILL.md files, or skill triggers.
Use when reviewing git diffs, commits, or pull requests for bugs, invariant violations, regressions, or security flaws.
Use when searching Amazon products, comparing live prices, filtering ratings, finding deals, or shortlisting items.
Use when querying Artificial Analysis or comparing AI model speed, latency, throughput, or provider token pricing.
Use when querying DeepSWE benchmark scores, leaderboards, pass@1 rates, model efficiency, or confidence intervals.
Use when asked to ship, get an issue or PR merge-ready or merged, or file a discovered problem for later.
Use when managing GitHub via gh CLI, issues, PRs, Actions, releases, or stacked PRs; not for contribution guidelines.
Use when asked to autommit, split changes into atomic commits, or create unattended git commits from local diffs.