Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
Use when the user wants to stress-test a plan against the project's existing domain language, glossary, code behavior, and documented decisions. Also use when they mention grilling a plan, sharpening terminology, walking a design tree, updating CONTEXT.md, or deciding whether an ADR is needed.
Use when the user wants to compact the current conversation into a handoff document for another agent or future session to continue the work. Also use when they mention handoff, transfer context, continue in a new session, or summarize the session for the next agent.
Use when the user provides a Figma node, screenshot, or visual reference and wants a narrow frontend change to match it. Also use for small UI copy, layout, spacing, state, or component adjustments driven by a concrete design reference.
Use when the user asks to push, upload, publish, or "залей" changes to main. Also use when they provide a GitHub remote and want the current repository committed and pushed.
Use when a change has become too large, the user worries about extra lines or overengineering, or asks to simplify, reduce scope, avoid unrelated changes, or keep only the necessary diff.
Use when the user asks why a bug happens, wants an explanation grounded in code, or says to check the real code/data path before fixing. Also use for missing data, inconsistent UI state, broken workflows, and suspected backend/frontend contract issues.