CodingCossack

Dispatches one subagent per independent domain to parallelize investigation/fixes. Use when you have 2+ unrelated failures (e.g., separate failing test files, subsystems, bugs) with no shared state or ordering dependencies.

Disciplined plan execution for implementation tasks. Use when executing a saved implementation plan, following step-by-step instructions from a plan document.

Git branch completion workflow. Use when implementation is complete, tests pass, and a feature branch needs to be integrated via merge, pull request, or cleanup.

Assesses and responds to incoming code review feedback on PRs (reviewer comments, requested changes), especially when suggestions are unclear, technically questionable, or scope-expanding. Use before implementing review suggestions to align on intent and keep changes minimal.

Use when you need to request a code review for a PR/MR and want a consistent review brief (context, scope, risk areas, test instructions, acceptance criteria) before merge.

Sequential subagent execution with two-stage review gates for implementation plans. Use when executing multi-task plans in current session, when tasks need fresh subagent context to avoid pollution, when formal review cycles (spec compliance then code quality) are required between tasks, or when you need diff-based validation of each task before proceeding.

Root cause analysis for debugging. Use when bugs, test failures, or unexpected behavior have non-obvious causes, or after multiple fix attempts have failed.

Red-green-refactor development methodology requiring verified test coverage. Use for feature implementation, bugfixes, refactoring, or any behavior changes where tests must prove correctness.

Implement, review, debug, and refactor TanStack Start React Server Components in React 19 apps. Use when tasks mention @tanstack/react-start/rsc, renderServerComponent, createCompositeComponent, CompositeComponent, renderToReadableStream, createFromReadableStream, createFromFetch, Composite Components, React Flight streams, loader or query owned RSC caching, router.invalidate, structuralSharing: false, selective SSR, stale names like renderRsc or .validator, or migration from Next App Router RSC patterns. Do not use for generic SSR or non-TanStack RSC frameworks except brief comparison.

Use when implementing, debugging, refactoring, testing, reviewing, or completing non-trivial changes to software — bugs, flaky tests, regressions, refactors, migrations, API or schema changes, auth, permissions, secrets, payments, webhooks, persistence, uploads, routing, exports, UI and user-visible surfaces, external integrations, dead-code or legacy removal, and long-running or resumed multi-step work. Do not use for pure explanation or Q&A, greenfield scaffolding with no existing behaviour at stake, prompt writing, dedicated security audits, or visual design.

Use when creating, rewriting, optimising, reviewing, or handing off prompts, system prompts, developer instructions, agent briefs, delegation prompts, task specifications, or reusable prompt templates for GPT-5.5 or GPT-5.6, especially when wording is procedural, repetitive, over-scaffolded, vague about success, or unclear about evidence, autonomy, tools, validation, output, or stopping.

Use for a site-specific GEO / AI-search audit or implementation task on a real URL or domain. Covers crawler access, renderability and indexability, answer-level citability, llms.txt, structured data and entity signals, platform-specific AI search issues, and report generation from those findings. Do not use for generic SEO chat, generic copywriting, or agency CRM / proposal work unless the user explicitly asks for those inside a GEO engagement.

Reviews React code with useEffect or useLayoutEffect. Replaces effects with derived state, event handlers, query APIs, key resets, useSyncExternalStore, or named hooks. Use when writing components, refactoring effects, or reviewing diffs.

Use when turning product notes, briefs, specifications, or mixed mission packs into prompts for Claude Opus, Fable, or another design-capable Claude model whose primary result is a visual artefact, frontend, redesign, prototype, design system, or image-to-code implementation.

Use when writing, rewriting, reviewing, shortening, or handing off prompts, system prompts, agent briefs, skills, CLAUDE.md files, or task specifications targeting Opus 5 or Fable 5 — including when a prompt was tuned for an older model, has grown repetitive or over-scaffolded, or is unclear about scope, autonomy, evidence, output length, or stopping. Not for prompts whose primary result is a visual artefact; writing-claude-design-prompts owns those.

Use when writing, rewriting, reviewing, shortening, or handing off prompts, system prompts, agent briefs, skills, CLAUDE.md files, or task specifications targeting Opus 5 or Fable 5 — including when a prompt was tuned for an older model, has grown repetitive or over-scaffolded, or is unclear about scope, autonomy, evidence, output length, or stopping.

Use when writing, reviewing, planning, or implementing anything a person will see or use, including pages, screens, search, listings, forms, emails, errors, empty states, onboarding, copy, and frontend behaviour, even when the task is framed as backend, routing, taxonomy, schema, policy, data, testing, or infrastructure work.