How do I install the Cook skill?

Run npx skillmds add majiayu000/cook-4 in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

This skill has not completed SkillMD's automated safety review yet. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Cook?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

majiayu000 (@majiayu000) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Cook by majiayu000 — Agent Skill

Name: Cook
Rating: 4.2 (5 reviews)
Author: majiayu000

Cook

(ePost) Use when user says "implement", "build", "add a feature", "cook", "make this work", or "continue the plan" — dispatches platform-aware feature implementation for web, iOS, Android, or backend

majiayu000 e235bba 2 files · 3.2 KB Updated Sep 9, 2026 567 repo stars

File contents

Cook — Unified Implementation Command

Implement features with automatic platform detection.

Step 0 — Active Plan Resolution (when args are empty)

If $ARGUMENTS is empty, resolve the active plan before asking the user for a task:

Run: node .claude/scripts/get-active-plan.cjs
If result ≠ none: read the plan's plan.md, identify the first phase with status: pending, and implement it — no need to ask the user what to do.
If result = none: scan plans/*/plan.md for the most recently created plan with status: pending (plans just created by /plan that haven't been activated yet). Sort by directory name descending; take the first match.
If still nothing: ask the user for a task description.

When a plan is found via step 3 (frontmatter scan), run node .claude/scripts/set-active-plan.cjs <plan-dir> to activate it before proceeding.

Step 1 — Flag Override

If $ARGUMENTS starts with --fast: skip auto-detection, load references/fast-mode.md and execute directly. Remaining args are the task description. If $ARGUMENTS starts with --parallel: skip auto-detection, load references/parallel-mode.md and execute directly. Remaining args are the task description. Otherwise: continue to Platform Detection.

Aspect Files

File	Purpose
`references/fast-mode.md`	Direct implementation — skip plan question, implement immediately
`references/parallel-mode.md`	Parallel implementation for multi-module features

Platform Detection

Detect platform per skill-discovery protocol.

Complexity → Variant

Single file or clear task → fast (skip plan question)
Multi-file, one module → fast
Multi-module or unknowns → parallel
Has existing plan in ./plans/ → follow plan
Plan with 3+ independent tasks → consider subagent-driven mode (see subagent-driven-development skill)

Execution

Route to the detected platform agent with feature description and platform context.

$ARGUMENTS

IMPORTANT: Analyze the skills catalog and activate the skills needed for the detected platform.

---name: cook-4description: (ePost) Use when user says "implement", "build", "add a feature", "cook", "make this work", or "continue the plan" — dispatches platform-aware feature implementation for web, iOS, Android, or backend--- # Cook — Unified Implementation Command Implement features with automatic platform detection. ## Step 0 — Active Plan Resolution (when args are empty) **If `$ARGUMENTS` is empty**, resolve the active plan before asking the user for a task: 1. Run: `node .claude/scripts/get-active-plan.cjs` 2. **If result ≠ `none`**: read the plan's `plan.md`, identify the first phase with `status: pending`, and implement it — no need to ask the user what to do. 3. **If result = `none`**: scan `plans/*/plan.md` for the most recently created plan with `status: pending` (plans just created by `/plan` that haven't been activated yet). Sort by directory name descending; take the first match. 4. **If still nothing**: ask the user for a task description. When a plan is found via step 3 (frontmatter scan), run `node .claude/scripts/set-active-plan.cjs <plan-dir>` to activate it before proceeding. ## Step 1 — Flag Override If `$ARGUMENTS` starts with `--fast`: skip auto-detection, load `references/fast-mode.md` and execute directly. Remaining args are the task description. If `$ARGUMENTS` starts with `--parallel`: skip auto-detection, load `references/parallel-mode.md` and execute directly. Remaining args are the task description. Otherwise: continue to Platform Detection. ## Aspect Files | File | Purpose | |------|---------| | `references/fast-mode.md` | Direct implementation — skip plan question, implement immediately | | `references/parallel-mode.md` | Parallel implementation for multi-module features | ## Platform Detection Detect platform per `skill-discovery` protocol. ## Complexity → Variant - Single file or clear task → fast (skip plan question) - Multi-file, one module → fast - Multi-module or unknowns → parallel - Has existing plan in ./plans/ → follow plan - Plan with 3+ independent tasks → consider subagent-driven mode (see `subagent-driven-development` skill) ## Execution Route to the detected platform agent with feature description and platform context. <feature>$ARGUMENTS</feature> **IMPORTANT:** Analyze the skills catalog and activate the skills needed for the detected platform.

Frequently asked questions