Parallel exploration of implementation approaches - implements multiple variants simultaneously and lets tests determine the winner
This skill should be used when moving from design to implementation. Triggers on "let's build", "implement this", "looks good let's code", "ready to implement". Presents options for parallel agent competition (cookoff), single subagent, or local implementation. Each agent creates own plan from shared design for genuine variation.
Scoring framework for test-kitchen cookoff and omakase-off. Invoked at Phase 4 to evaluate implementations using 5-criteria scoring. Do not invoke directly - called by cookoff/omakase-off.
This skill should be used as the entry gate for build/create/implement requests. Triggers on "build X", "create Y", "implement Z", "add feature", "try both approaches", "not sure which approach". Offers brainstorm-together or omakase (chef's choice parallel exploration) options. Detects indecision during brainstorming to offer parallel exploration.
You have a spec. Three valid approaches exist. Instead of guessing which one to build, Test Kitchen builds all of them in parallel git worktrees and lets your tests pick the winner.
Read more: Cookoff: Same Spec, Different Code · Omakase: Show Me
Parallel implementation framework with two gate skills:
| Skill | Gate | Trigger |
|---|---|---|
test-kitchen:omakase-off | Entry | FIRST on any build/create/implement request |
test-kitchen:cookoff | Exit | At design→implementation transition |
/plugin marketplace add 2389-research/claude-plugins
/plugin install test-kitchen@2389-research
"Build X" / "Create Y" / "Implement Z"
↓
┌─────────────────────────────────────┐
│ OMAKASE-OFF (entry gate) │
│ Wraps brainstorming │
│ │
│ Choice: │
│ 1. Brainstorm together │
│ 2. Omakase (3-5 parallel designs) │
└─────────────────────────────────────┘
↓
[Brainstorming / Design phase]
↓
Design complete, "let's implement"
↓
┌─────────────────────────────────────┐
│ COOKOFF (exit gate) │
│ Wraps implementation │
│ │
│ Choice: │
│ 1. Cookoff (2-5 parallel agents) │
│ 2. Single subagent │
│ 3. Local implementation │
└─────────────────────────────────────┘
↓
[Implementation]
When you make a build/create request:
User: "Build a notification system"
Claude: I'm using test-kitchen:omakase-off.
How would you like to explore this?
1. Brainstorm together - We'll work through the design step by step
2. Omakase - I'll generate 3-5 best approaches, implement in parallel, tests pick winner
Which approach?
Option 1 (Brainstorm) proceeds to brainstorming skill (or fallback brainstorming), with passive slot detection.
Option 2 (Omakase) does the following:
During brainstorming, slot detection can also trigger it:
Claude: What storage approach would you prefer?
User: "not sure, either could work"
Claude: [marks storage as architectural slot]
Claude: For authentication?
User: "no preference, you pick"
Claude: [marks auth as slot, detects 2+ uncertain responses]
You seem flexible on the approach. Would you like to:
1. I'll pick what seems best and continue
2. Explore multiple approaches in parallel (omakase-off)
→ I'll implement 2-3 variants and let tests decide
User: "Looks good, let's implement"
Claude: I'm using test-kitchen:cookoff.
How would you like to implement this design?
1. Cookoff (recommended) - 3 parallel agents, each creates own plan, pick best
→ Complexity: medium feature
2. Single subagent - One agent plans and implements
3. Local - Plan and implement here
Which approach?
Option 1 (Cookoff):
Options 2/3 -- single agent or local implementation proceeds normally.
Skills can't passively detect "uncertainty" or "readiness" -- they must claim specific moments in the conversation flow.
Test Kitchen orchestrates these skills (falls back gracefully if not installed):
superpowers:brainstormingsuperpowers:writing-planssuperpowers:executing-planssuperpowers:using-git-worktreessuperpowers:dispatching-parallel-agentssuperpowers:test-driven-developmentsuperpowers:verification-before-completionscenario-testing:skillsfresh-eyes-review:skillssuperpowers:finishing-a-development-branchNamed after a restaurant test kitchen, where chefs try multiple approaches before putting a dish on the menu.
If Test Kitchen saved you from building the wrong approach first, a ⭐ helps us know it's landing.
Built by 2389 · Part of the Claude Code plugin marketplace
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
npx claudepluginhub 2389-research/claude-plugins --plugin test-kitchenMCP server for agent social media platform - enables AI agents to interact in team-based discussions
Iterative artifact refinement with investigation-first judge board - constructs problem-specific judges that read the code, understand the problem, and propose evidence-based improvements
Tiered-delegation execution where Sonnet plans a spec into sprints, a cheap Haiku agent builds and self-verifies against the gate, and a scoped Sonnet fix runs only on failure - benchmarked ~64% cheaper than Opus at equal quality.
Agentic binary reverse engineering for ELF binaries on ARM64, ARMv7, x86_64 - hypothesis-driven analysis with radare2, Ghidra, GDB, QEMU
A comprehensive MCP server that provides Claude with private journaling, semantic search, resources, and guided prompts for reflection and learning
Thinking partner for design decisions and architecture brainstorming. Transforms Claude from a solution-provider into a senior engineer thinking partner.
This skill should be used when users need to generate ideas, explore creative solutions, or systematically brainstorm approaches to problems. Use when users request help with ideation, content planning, product features, marketing campaigns, strategic planning, creative writing, or any task requiring structured idea generation. The skill provides 30+ research-validated prompt patterns across 14 categories with exact templates, success metrics, and domain-specific applications.
Brainstorm, plan, debug, review, and compound learnings with AI agents
複数の実装アプローチを並列チームで設計・議論し、比較して最適な計画を選ぶスキル。各アプローチに専門グループ(architect, critic, risk-analyst)が付き、グループ内で議論して計画を洗練する。
Claude plans, delegates, and tastes every plate. Sonnet, Codex, or GLM does the knife work. A Claude Code plugin for the orchestrator/implementer split: /serve runs the whole line autonomously (implement, cross-review, fix, verify), /fire hands specced implementation to the worker, /taste runs cross-model review, /refire fixes what the review found, /simmer loops the worker under Claude's judgment, /mise sets up the kitchen, /receipts prints the check.
Per-feature development lifecycle: design, launch, build, wrap.