원클릭으로 Manus에서 모든 스킬 실행

시작하기

hi-cook

스타0

포크0

업데이트2026년 6월 11일 03:42

ALWAYS activate before implementing ANY feature, plan, or fix.

설치

Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.

Manus에서 실행

출처

baka3k

baka3k/agent-skills

GitHub 저장소 열기 Creator 저장소 보기

다운로드

Manus에서 실행

Cook - Feature Implementation

Do NOT write code until a plan exists and has been reviewed. User override: "just code it" or "skip planning" - then respect.

Intent Detection

Input	Mode	Behavior
--full, "full"	full	Full workflow với research + review
--review	review	Có review gate ở cuối
--auto, "trust me", "yolo"	auto	Auto-approve all
--no-test	no-test	Skip testing
Path to plan.md/phase-*.md	code	Execute existing plan
Default	fast	Skip research, skip review, test nhanh

Process Flow

[Plan] -> [Implement] -> [Test] -> [Finalize]

Mode Matrix

Mode	Research	Review	Testing
fast (default)	Skip	Skip	Run command
full	Yes	MUST	Run command
review	Skip	MUST	Run command
auto	Skip	Auto-pass	Run command
no-test	Skip	Skip	Skip

Steps

Step 1: Plan (skip if plan.md provided)

Dùng sequential-thinking để phân tích task ngắn gọn
Dùng docs-seeker nếu cần tra cứu tài liệu
Không spawn planner riêng. Gọi hi:plan --fast inline nếu cần.

Step 2: Implement

Execute tasks trực tiếp. TaskUpdate in_progress trên mỗi task. Parallel mode: launch fullstack-developer per phase.

Step 3: Test (skip no-test)

Run test command. Xem output. Nếu fail:

Lần 1-2: tự phân tích và fix
Lần 3+: spawn hi:fix để debug chuyên sâu Không spawn tester riêng.

Step 4: Finalize

TaskUpdate all tasks complete
git commit
/hi:journal

Review (optional, flag --review hoặc full mode)

Run code-reviewer. Score>=9.5 + 0 critical = auto-approve (auto mode only). Max 3 fix cycles. Critical issues always block: Security, Performance, Architecture violations.

이 저장소의 다른 Skills

같은 저장소

ck-docs-seeker

baka3k/agent-skills

Search library/framework documentation via llms.txt (context7.com). Use for API docs, GitHub repository analysis, technical documentation lookup, latest library features.

2026-06-110

ck-debug

baka3k/agent-skills

Debug systematically with root cause analysis before fixes. Use for bugs, test failures, unexpected behavior, performance issues, call stack tracing, multi-layer validation, log analysis, CI/CD failures, database diagnostics, system investigation.

2026-06-110

hi-fix

baka3k/agent-skills

ALWAYS activate before fixing ANY bug, error, test failure, CI/CD issue, type error, lint, log error, UI issue, code problem.

2026-06-110

knows

baka3k/agent-skills

Unified knowledge retrieval skill. Build traceable answers from git, MCP, and memory files.

2026-06-110

hi-plan

baka3k/agent-skills

Plan implementations, design architectures, create technical roadmaps with detailed phases.

2026-06-110

ck-problem-solving

baka3k/agent-skills

Apply systematic problem-solving techniques when stuck. Use for complexity spirals, innovation blocks, recurring patterns, assumption constraints, simplification cascades, scale uncertainty.

2026-06-110

name	hi:cook
description	ALWAYS activate before implementing ANY feature, plan, or fix.
argument-hint	[task] [--full\|--review\|--auto\|--no-test] — default: fast mode
metadata	{"author":"baka3k","version":"3.0.0"}

Cook - Feature Implementation

Do NOT write code until a plan exists and has been reviewed. User override: "just code it" or "skip planning" - then respect.

Intent Detection

Input	Mode	Behavior
--full, "full"	full	Full workflow với research + review
--review	review	Có review gate ở cuối
--auto, "trust me", "yolo"	auto	Auto-approve all
--no-test	no-test	Skip testing
Path to plan.md/phase-*.md	code	Execute existing plan
Default	fast	Skip research, skip review, test nhanh

Process Flow

[Plan] -> [Implement] -> [Test] -> [Finalize]

Mode Matrix

Mode	Research	Review	Testing
fast (default)	Skip	Skip	Run command
full	Yes	MUST	Run command
review	Skip	MUST	Run command
auto	Skip	Auto-pass	Run command
no-test	Skip	Skip	Skip

Steps

Step 1: Plan (skip if plan.md provided)

Dùng sequential-thinking để phân tích task ngắn gọn
Dùng docs-seeker nếu cần tra cứu tài liệu
Không spawn planner riêng. Gọi hi:plan --fast inline nếu cần.

Step 2: Implement

Execute tasks trực tiếp. TaskUpdate in_progress trên mỗi task. Parallel mode: launch fullstack-developer per phase.

Step 3: Test (skip no-test)

Run test command. Xem output. Nếu fail:

Lần 1-2: tự phân tích và fix
Lần 3+: spawn hi:fix để debug chuyên sâu Không spawn tester riêng.

Step 4: Finalize

TaskUpdate all tasks complete
git commit
/hi:journal

Review (optional, flag --review hoặc full mode)

Run code-reviewer. Score>=9.5 + 0 critical = auto-approve (auto mode only). Max 3 fix cycles. Critical issues always block: Security, Performance, Architecture violations.