菜单
Implement features end-to-end with smart intent detection. Accepts natural language or plan paths. Auto-routes to appropriate workflow (fast, auto, parallel, code).
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
Build high-converting SaaS affiliate programs with 20-40% commissions, KOL/KOC partnerships, and fraud prevention. Covers platform selection (PartnerStack, FirstPromoter, Rewardful), commission structures (recurring vs one-time, tiered), influencer outreach strategies, FTC/GDPR compliance, risk management, and case studies (Dropbox 3900%, PayPal 100M users). Use for designing affiliate programs, recruiting partners, optimizing conversion rates, preventing fraud, or scaling referral revenue.
Prompt + image gen. 6000+ examples. Dual-option: Standard (Flash) or Creative (Pro). Generate infographics, thumbnails, avatars, product shots with Nano Banana.
Tạo một Claude skill mới theo đúng chuẩn Agent Skills. Dùng khi người dùng muốn tạo skill, tạo claude skill, hoặc thêm khả năng mới cho Claude trong project này.
Git operations with conventional commits. Use for staging, committing, pushing, PRs, merges. Auto-splits commits by type/scope. Security scans for secrets.
Fast codebase scouting using parallel agents. Use for file discovery, task context gathering, quick searches across directories. Supports internal (Explore) and external (Gemini/OpenCode) agents.
Tạo bộ kịch bản video TikTok affiliate hoàn chỉnh (5 video) cho một sản phẩm. Dùng khi cần viết script, kịch bản, lời thoại cho video TikTok bán hàng affiliate.
| name | cook |
| description | Implement features end-to-end with smart intent detection. Accepts natural language or plan paths. Auto-routes to appropriate workflow (fast, auto, parallel, code). |
| version | 2.0.0 |
End-to-end implementation with automatic workflow detection.
Principles: YAGNI, KISS, DRY | Token efficiency | Concise reports
/cook <natural language task OR plan path>
Optional flags: --fast, --parallel, --no-test, --auto
| Input Pattern | Detected Mode | Behavior |
|---|---|---|
Path to plan.md or phase-*.md | code | Execute existing plan |
| Contains "fast", "quick" | fast | Skip research, scout→plan→code |
| Contains "trust me", "auto" | auto | Auto-approve all steps |
| Lists 3+ features OR "parallel" | parallel | Multi-agent execution |
| Contains "no test", "skip test" | no-test | Skip testing step |
| Default | interactive | Full workflow with user input |
See references/intent-detection.md for detection logic.
[Intent Detection] → [Research?] → [Plan] → [Implement] → [Test?] → [Review] → [Finalize]
| Mode | Research | Testing | Review | All Phases |
|---|---|---|---|---|
| interactive | ✓ | ✓ | User approval | One at a time |
| auto | ✓ | ✓ | Auto if score≥9.5 | All at once |
| fast | ✗ | ✓ | Simplified | One at a time |
| parallel | Optional | ✓ | User approval | Parallel groups |
| no-test | ✓ | ✗ | User approval | One at a time |
| code | ✗ | ✓ | User approval | Per plan |
✓ Step [N]: [Brief status] - [Key metrics]
| Phase | Subagent |
|---|---|
| Research | researcher (parallel, optional in fast) |
| Scout | scout |
| Plan | planner |
| UI Work | ui-ux-designer |
| Testing | tester, debugger |
| Review | code-reviewer |
| Finalize | project-manager, docs-manager, git-manager |
references/intent-detection.md - Detection rules and routing logicreferences/workflow-steps.md - Detailed step definitions for all modesreferences/review-cycle.md - Interactive and auto review processesreferences/subagent-patterns.md - Subagent invocation patterns