@mrtrinhvn/ag-kit 1.0.11 → 1.1.1

package/.agent/skills/brainstorming/dynamic-questioning.md

@@ -1,350 +0,0 @@
-# Dynamic Question Generation
-
-> **PRINCIPLE:** Questions are not about gathering data—they are about **revealing architectural consequences**.
->
-> Every question must connect to a concrete implementation decision that affects cost, complexity, or timeline.
-
----
-
-## 🧠 Core Principles
-
-### 1. Questions Reveal Consequences
-
-A good question is not "What color do you want?" but:
-
-```markdown
-❌ BAD: "What authentication method?"
-✅ GOOD: "Should users sign up with email/password or social login?
-
-   Impact:
-   - Email/Pass → Need password reset, hashing, 2FA infrastructure
-   - Social → OAuth providers, user profile mapping, less control
-
-   Trade-off: Security vs. Development time vs. User friction"
-```
-
-### 2. Context Before Content
-
-First understand **where** this request fits:
-
-| Context | Question Focus |
-|---------|----------------|
-| **Greenfield** (new project) | Foundation decisions: stack, hosting, scale |
-| **Feature Addition** | Integration points, existing patterns, breaking changes |
-| **Refactor** | Why refactor? Performance? Maintainability? What's broken? |
-| **Debug** | Symptoms → Root cause → Reproduction path |
-
-### 3. Minimum Viable Questions
-
-**PRINCIPLE:** Each question must eliminate a fork in the implementation road.
-
-```
-Before Question:
-├── Path A: Do X (5 min)
-├── Path B: Do Y (15 min)
-└── Path C: Do Z (1 hour)
-
-After Question:
-└── Path Confirmed: Do X (5 min)
-```
-
-If a question doesn't reduce implementation paths → **DELETE IT**.
-
-### 4. Questions Generate Data, Not Assumptions
-
-```markdown
-❌ ASSUMPTION: "User probably wants Stripe for payments"
-✅ QUESTION: "Which payment provider fits your needs?
-
-   Stripe → Best documentation, 2.9% + $0.30, US-centric
-   LemonSqueezy → Merchant of Record, 5% + $0.50, global taxes
-   Paddle → Complex pricing, handles EU VAT, enterprise focus"
-```
-
----
-
-## 📋 Question Generation Algorithm
-
-```
-INPUT: User request + Context (greenfield/feature/refactor/debug)
-│
-├── STEP 1: Parse Request
-│   ├── Extract domain (ecommerce, auth, realtime, cms, etc.)
-│   ├── Extract features (explicit and implied)
-│   └── Extract scale indicators (users, data volume, frequency)
-│
-├── STEP 2: Identify Decision Points
-│   ├── What MUST be decided before coding? (blocking)
-│   ├── What COULD be decided later? (deferable)
-│   └── What has ARCHITECTURAL impact? (high-leverage)
-│
-├── STEP 3: Generate Questions (Priority Order)
-│   ├── P0: Blocking decisions (cannot proceed without answer)
-│   ├── P1: High-leverage (affects >30% of implementation)
-│   ├── P2: Medium-leverage (affects specific features)
-│   └── P3: Nice-to-have (edge cases, optimization)
-│
-└── STEP 4: Format Each Question
-    ├── What: Clear question
-    ├── Why: Impact on implementation
-    ├── Options: Trade-offs (not just A vs B)
-    └── Default: What happens if user doesn't answer
-```
-
----
-
-## 🎯 Domain-Specific Question Banks
-
-### E-Commerce
-
-| Question | Why It Matters | Trade-offs |
-|----------|----------------|------------|
-| **Single or Multi-vendor?** | Multi-vendor → Commission logic, vendor dashboards, split payments | +Revenue, -Complexity |
-| **Inventory Tracking?** | Needs stock tables, reservation logic, low-stock alerts | +Accuracy, -Development time |
-| **Digital or Physical Products?** | Digital → Download links, no shipping | Physical → Shipping APIs, tracking |
-| **Subscription or One-time?** | Subscription → Recurring billing, dunning, proration | +Revenue, -Complexity |
-
-### Authentication
-
-| Question | Why It Matters | Trade-offs |
-|----------|----------------|------------|
-| **Social Login Needed?** | OAuth providers vs. password reset infrastructure | +UX, -Control |
-| **Role-Based Permissions?** | RBAC tables, policy enforcement, admin UI | +Security, -Development time |
-| **2FA Required?** | TOTP/SMI infrastructure, backup codes, recovery flow | +Security, -UX friction |
-| **Email Verification?** | Verification tokens, email service, resend logic | +Security, -Sign-up friction |
-
-### Real-time
-
-| Question | Why It Matters | Trade-offs |
-|----------|----------------|------------|
-| **WebSocket or Polling?** | WS → Server scaling, connection management | Polling → Simpler, higher latency |
-| **Expected Concurrent Users?** | <100 → Single server, >1000 → Redis pub/sub, >10k → specialized infra | +Scale, -Complexity |
-| **Message Persistence?** | History tables, storage costs, pagination | +UX, -Storage |
-| **Ephemeral or Durable?** | Ephemeral → In-memory, Durable → Database write before emit | +Reliability, -Latency |
-
-### Content/CMS
-
-| Question | Why It Matters | Trade-offs |
-|----------|----------------|------------|
-| **Rich Text or Markdown?** | Rich Text → Sanitization, XSS risks | Markdown → Simple, no WYSIWYG |
-| **Draft/Publish Workflow?** | Status field, scheduled jobs, versioning | +Control, -Complexity |
-| **Media Handling?** | Upload endpoints, storage, optimization | +Features, -Development time |
-| **Multi-language?** | i18n tables, translation UI, fallback logic | +Reach, -Complexity |
-
----
-
-## 📐 Dynamic Question Template
-
-```markdown
-Based on your request for [DOMAIN] [FEATURE]:
-
-## 🔴 CRITICAL (Blocking Decisions)
-
-### 1. **[DECISION POINT]**
-
-**Question:** [Clear, specific question]
-
-**Why This Matters:**
-- [Explain architectural consequence]
-- [Affects: cost / complexity / timeline / scale]
-
-**Options:**
-| Option | Pros | Cons | Best For |
-|--------|------|------|----------|
-| A | [Advantage] | [Disadvantage] | [Use case] |
-| B | [Advantage] | [Disadvantage] | [Use case] |
-
-**If Not Specified:** [Default choice + rationale]
-
----
-
-## 🟡 HIGH-LEVERAGE (Affects Implementation)
-
-### 2. **[DECISION POINT]**
-[Same format]
-
----
-
-## 🟢 NICE-TO-HAVE (Edge Cases)
-
-### 3. **[DECISION POINT]**
-[Same format]
-```
-
----
-
-## 🔄 Iterative Questioning
-
-### First Pass (3-5 Questions)
-Focus on **blocking decisions**. Don't proceed without answers.
-
-### Second Pass (After Initial Implementation)
-As patterns emerge, ask:
-- "This feature implies [X]. Should we handle [edge case] now or defer?"
-- "We're using [Pattern A]. Should [Feature B] follow the same pattern?"
-
-### Third Pass (Optimization)
-When functionality works:
-- "Performance bottleneck at [X]. Optimize now or acceptable for now?"
-- "Refactor [Y] for maintainability or ship as-is?"
-
----
-
-## 🎭 Example: Full Question Generation
-
-```
-USER REQUEST: "Build an Instagram clone"
-
-STEP 1: Parse
-├── Domain: Social Media
-├── Features: Photo sharing, engagement (likes/comments), user profiles
-├── Implied: Feed, following, authentication
-└── Scale: Potentially high (social apps go viral)
-
-STEP 2: Decision Points
-├── Blocking: Storage strategy, authentication method, feed type
-├── High-leverage: Real-time notifications, data model complexity
-└── Deferable: Analytics, advanced search, reels/video
-
-STEP 3: Generate Questions (Priority)
-
-P0 (Blocking):
-1. Storage Strategy → Affects architecture, cost, speed
-2. Feed Algorithm → Affects database queries, complexity
-3. Auth Method → Affects dev time, UX, security
-
-P1 (High-leverage):
-4. Real-time Notifications → WebSocket vs polling
-5. Media Processing → Client-side vs server-side optimization
-
-P2 (Deferable):
-6. Story/Reels → Major feature creep, defer to v2
-7. DM/Chat → Separate subsystem, defer to v2
-
-STEP 4: Format Output
-```
-
----
-
-## 📊 Generated Output (Example)
-
-```
-Based on your Instagram clone request:
-
-## 🔴 CRITICAL DECISIONS (Cannot Proceed Without Answers)
-
-### 1. **Photo Storage Strategy**
-
-**Question:** Where will user photos be stored and served?
-
-**Why This Matters:**
-- Affects: Monthly hosting costs, page load speed, CDN complexity
-- High-volume social apps: 1000 users × 10 photos × 2MB = 20GB storage
-
-**Options:**
-| Option | Cost | Speed | Complexity | Best For |
-|--------|------|-------|------------|----------|
-| **Cloudinary** | $89/mo (25GB) | Fast (CDN) | Low | MVP, rapid launch |
-| **AWS S3 + CloudFront** | $0.023/GB | Fast (CDN) | Medium | Production, cost-optimized |
-| **Supabase Storage** | Free tier 1GB | Medium | Low | Small scale, simple |
-| **Local Storage** | Server cost | Slow | Low | Development only |
-
-**If Not Specified:** Cloudinary (balanced for MVP)
-
----
-
-### 2. **Feed Algorithm Choice**
-
-**Question:** How should the main feed work?
-
-**Why This Matters:**
-- Affects: Database query complexity, caching strategy, development time
-- Algorithmic feeds require ML infrastructure, ranking signals
-
-**Options:**
-| Option | Complexity | Database Impact | Best For |
-|--------|------------|-----------------|----------|
-| **Chronological** | Low | Simple query | Early stage, transparency |
-| **Follow-Only** | Medium | JOIN with pagination | Most social apps |
-| **Algorithmic** | High | Pre-computed feed tables | Instagram competitor |
-
-**If Not Specified:** Follow-Only (balanced for social app)
-
----
-
-### 3. **Authentication Approach**
-
-**Question:** How do users sign up and login?
-
-**Why This Matters:**
-- Affects: Development time (2-5 hours), security posture, UX friction
-
-**Options:**
-| Option | Dev Time | Security | UX | Best For |
-|--------|----------|----------|-----|----------|
-| **Email/Password** | 4-5 hrs | High (if 2FA) | Medium | Full control needed |
-| **Social Only** | 1-2 hrs | Provider-dependent | Smooth | B2C, rapid launch |
-| **Magic Link** | 2-3 hrs | Medium | Very smooth | Security-focused |
-| **Clerk/Auth0** | 1 hr | High | Smooth | Fastest to market |
-
-**If Not Specified:** Clerk (fastest for MVP)
-
----
-
-## 🟡 HIGH-LEVERAGE (Affects Architecture)
-
-### 4. **Real-time Notifications**
-
-**Question:** Do users need instant notifications for likes/comments?
-
-**Why This Matters:**
-- WebSocket adds infrastructure complexity (Redis pub/sub for scaling)
-- Polling is simpler but higher latency
-
-**Options:**
-| Option | Complexity | Scale Cost | Best For |
-|--------|------------|------------|----------|
-| **WebSocket + Redis** | High | $10+/mo | >1000 concurrent users |
-| **Polling (30s)** | Low | DB queries | <1000 users |
-| **No Real-time** | None | None | MVP, validate first |
-
-**If Not Specified:** Polling for MVP (defer WebSocket until validated)
-
----
-
-## 🟢 NICE-TO-HAVE (Defer to v2)
-
-### 5. **Video/Reels Support**
-- Major complexity (video processing, streaming infrastructure)
-- Recommendation: Launch with photos only, add video after validation
-
-### 6. **Direct Messaging**
-- Separate subsystem (chat infrastructure different from feed)
-- Recommendation: Use Pusher/Stream for real-time or defer entirely
-
----
-
-## 📋 Summary
-
-| Decision | Recommendation | If Changed |
-|----------|----------------|------------|
-| Storage | Cloudinary | +3 hrs setup |
-| Feed | Follow-only | +2 hrs query optimization |
-| Auth | Clerk | -3 hrs dev time |
-| Real-time | Polling | +5 hrs WebSocket setup |
-| Video | Defer to v2 | N/A |
-| DM | Defer to v2 | N/A |
-
-**Total Estimated MVP Time:** 15-20 hours with recommendations above
-```
-
----
-
-## 🎯 Principles Recap
-
-1. **Every question = Architectural decision** → Not data gathering
-2. **Show trade-offs** → User understands consequences
-3. **Prioritize blocking decisions** → Cannot proceed without
-4. **Provide defaults** → If user doesn't answer, we proceed anyway
-5. **Domain-aware** → Ecommerce questions ≠ Auth questions ≠ Real-time questions
-6. **Iterative** → More questions as patterns emerge during implementation

package/.agent/skills/business-ops/SKILL.md

@@ -1,26 +0,0 @@
----
-name: business-ops
-description: Skill Quản trị Vận Hành. Đọc soát Hợp đồng (Legal), Phân tích Kế toán (Finance), và Quản lý Tuyển dụng (HR).
-allowed-tools: Read, Bash
-version: 1.0
-priority: MODERATE
----
-
-# Kỹ năng Vận Hành Doanh Nghiệp (Business Operations)
-
-> **MỤC TIÊU:** Hỗ trợ Giám đốc xử lý các mảng "Back-Office" ngoài luồng lập trình để Tiết kiệm thời gian quản trị.
-
-Các phân hệ Vận hành cụ thể:
-
-## 1. Finance Ops (Nghiệp vụ Tài chính - Kế toán)
-*   Kiểm tra Bảng giá API các dịch vụ Sub (AWS, Supabase, Vercel) xem code có vi phạm rủi ro "Tràn hóa đơn" (Billing Attack) không?
-*   Gợi ý giải pháp tiết kiệm Cost Token khi gọi LLM Agent.
-
-## 2. Legal & Compliance (Pháp lý - Hợp đồng)
-*   Đọc hiểu Terms of Service (TOS) của các Provider (như DNSE, Yuanta, SSI).
-*   Cảnh báo rủi ro về Privacy (Lưu Mật khẩu User dạng Plain text là vi phạm GDPR). 
-*   Quét Compliance: Đảm bảo không lộ Key, không vi phạm quy chuẩn chia sẻ dữ liệu chứng khoán.
-
-## 3. HR & Ops (Tuyển mộ - Quy trình)
-*   User có thể yêu cầu: "Viết giúp JD để tuyển 3 ông lính ReactJS". Hệ thống sẽ dùng kiến thức Codebase hiện tại để sinh ra 1 bản Yêu cầu Công việc sát thực tế nhất 100%. (Bắt buộc phải biết Redux? Hook? Tailwind? Hệ thống tự đọc Code của dự án và sinh ra luật).
-*   Tạo Quy trình làm việc (Standard Operating Procedure - SOP) cho Developer mới Onboard.

package/.agent/skills/clean-code/SKILL.md

@@ -1,202 +0,0 @@
----
-name: clean-code
-description: Pragmatic coding standards - concise, direct, no over-engineering, no unnecessary comments
-allowed-tools: Read, Write, Edit
-version: 2.0
-priority: CRITICAL
----
-
-# Clean Code - Pragmatic AI Coding Standards
-
-> **CRITICAL SKILL** - Be **concise, direct, and solution-focused**.
-
----
-
-## Core Principles
-
-| Principle | Rule |
-|-----------|------|
-| **SRP** | Single Responsibility - each function/class does ONE thing |
-| **DRY** | Don't Repeat Yourself - extract duplicates, reuse |
-| **KISS** | Keep It Simple - simplest solution that works |
-| **YAGNI** | You Aren't Gonna Need It - don't build unused features |
-| **Boy Scout** | Leave code cleaner than you found it |
-
----
-
-## Naming Rules
-
-| Element | Convention |
-|---------|------------|
-| **Variables** | Reveal intent: `userCount` not `n` |
-| **Functions** | Verb + noun: `getUserById()` not `user()` |
-| **Booleans** | Question form: `isActive`, `hasPermission`, `canEdit` |
-| **Constants** | SCREAMING_SNAKE: `MAX_RETRY_COUNT` |
-
-> **Rule:** If you need a comment to explain a name, rename it.
-
----
-
-## Function Rules
-
-| Rule | Description |
-|------|-------------|
-| **Small** | Max 20 lines, ideally 5-10 |
-| **One Thing** | Does one thing, does it well |
-| **One Level** | One level of abstraction per function |
-| **Few Args** | Max 3 arguments, prefer 0-2 |
-| **No Side Effects** | Don't mutate inputs unexpectedly |
-
----
-
-## Code Structure
-
-| Pattern | Apply |
-|---------|-------|
-| **Guard Clauses** | Early returns for edge cases |
-| **Flat > Nested** | Avoid deep nesting (max 2 levels) |
-| **Composition** | Small functions composed together |
-| **Colocation** | Keep related code close |
-
----
-
-## AI Coding Style
-
-| Situation | Action |
-|-----------|--------|
-| User asks for feature | STOP. Use Socratic Gate (brainstorming) |
-| User reports bug | Fix it, don't explain |
-| No clear requirement | Ask, don't assume |
-| Need current time/date | ALWAYS use `get_time_machine().now()` or `.is_trading_hours()`. NEVER use `datetime.now()` directly. |
-
----
-
-## Anti-Patterns (DON'T)
-
-| ❌ Pattern | ✅ Fix |
-|-----------|-------|
-| Comment every line | Delete obvious comments |
-| Helper for one-liner | Inline the code |
-| Factory for 2 objects | Direct instantiation |
-| utils.ts with 1 function | Put code where used |
-| "First we import..." | Just write code |
-| Deep nesting | Guard clauses |
-| Magic numbers | Named constants |
-| God functions | Split by responsibility |
-
----
-
-## 🔴 Before Editing ANY File (THINK FIRST!)
-
-**Before changing a file, ask yourself:**
-
-| Question | Why |
-|----------|-----|
-| **What imports this file?** | They might break |
-| **What does this file import?** | Interface changes |
-| **What tests cover this?** | Tests might fail |
-| **Is this a shared component?** | Multiple places affected |
-
-**Quick Check:**
-```
-File to edit: UserService.ts
-└── Who imports this? → UserController.ts, AuthController.ts
-└── Do they need changes too? → Check function signatures
-```
-
-> 🔴 **Rule:** Edit dependent files using independent 5-minute micro-tasks strictly within the Plan.
-> 🔴 **Never leave broken imports or missing updates.**
-
----
-
-## Summary
-
-| Do | Don't |
-|----|-------|
-| Ask questions before big features | Write tutorials |
-| Let code self-document | Add obvious comments |
-| Fix bugs immediately | Explain the fix first |
-| Inline small things | Create unnecessary files |
-| Name things clearly | Use abbreviations |
-| Keep functions small | Write 100+ line functions |
-
-> **Remember: The user wants working code, not a programming lesson.**
-
----
-
-## 🔴 Self-Check Before Completing (MANDATORY)
-
-**Before saying "task complete", verify:**
-
-| Check | Question |
-|-------|----------|
-| ✅ **Goal met?** | Did I do exactly what user asked? |
-| ✅ **Files edited?** | Did I modify all necessary files? |
-| ✅ **Code works?** | Did I test/verify the change? |
-| ✅ **No errors?** | Lint and TypeScript pass? |
-| ✅ **Nothing forgotten?** | Any edge cases missed? |
-
-> 🔴 **Rule:** If ANY check fails, fix it before completing.
-
----
-
-## Verification Scripts (MANDATORY)
-
-> 🔴 **CRITICAL:** Each agent runs ONLY their own skill's scripts after completing work.
-
-### Agent → Script Mapping
-
-| Agent | Script | Command |
-|-------|--------|---------|
-| **frontend-specialist** | UX Audit | `python .agent/skills/frontend-design/scripts/ux_audit.py .` |
-| **frontend-specialist** | A11y Check | `python .agent/skills/frontend-design/scripts/accessibility_checker.py .` |
-| **backend-specialist** | API Validator | `python .agent/skills/api-patterns/scripts/api_validator.py .` |
-| **mobile-developer** | Mobile Audit | `python .agent/skills/mobile-design/scripts/mobile_audit.py .` |
-| **database-architect** | Schema Validate | `python .agent/skills/database-design/scripts/schema_validator.py .` |
-| **security-auditor** | Security Scan | `python .agent/skills/vulnerability-scanner/scripts/security_scan.py .` |
-| **seo-specialist** | SEO Check | `python .agent/skills/seo-fundamentals/scripts/seo_checker.py .` |
-| **seo-specialist** | GEO Check | `python .agent/skills/geo-fundamentals/scripts/geo_checker.py .` |
-| **performance-optimizer** | Lighthouse | `python .agent/skills/performance-profiling/scripts/lighthouse_audit.py <url>` |
-| **test-engineer** | Test Runner | `python .agent/skills/testing-patterns/scripts/test_runner.py .` |
-| **test-engineer** | Playwright | `python .agent/skills/webapp-testing/scripts/playwright_runner.py <url>` |
-| **Any agent** | Lint Check | `python .agent/skills/lint-and-validate/scripts/lint_runner.py .` |
-| **Any agent** | Type Coverage | `python .agent/skills/lint-and-validate/scripts/type_coverage.py .` |
-| **Any agent** | i18n Check | `python .agent/skills/i18n-localization/scripts/i18n_checker.py .` |
-
-> ❌ **WRONG:** `test-engineer` running `ux_audit.py`
-> ✅ **CORRECT:** `frontend-specialist` running `ux_audit.py`
-
----
-
-### 🔴 Script Output Handling (READ → SUMMARIZE → ASK)
-
-**When running a validation script, you MUST:**
-
-1. **Run the script** and capture ALL output
-2. **Parse the output** - identify errors, warnings, and passes
-3. **Summarize to user** in this format:
-
-```markdown
-## Script Results: [script_name.py]
-
-### ❌ Errors Found (X items)
-- [File:Line] Error description 1
-- [File:Line] Error description 2
-
-### ⚠️ Warnings (Y items)
-- [File:Line] Warning description
-
-### ✅ Passed (Z items)
-- Check 1 passed
-- Check 2 passed
-
-**Should I fix the X errors?**
-```
-
-4. **Wait for user confirmation** before fixing
-5. **After fixing** → Re-run script to confirm
-
-> 🔴 **VIOLATION:** Running script and ignoring output = FAILED task.
-> 🔴 **VIOLATION:** Auto-fixing without asking = Not allowed.
-> 🔴 **Rule:** Always READ output → SUMMARIZE → ASK → then fix.
-

package/.agent/skills/cli-generator/SKILL.md

@@ -1,48 +0,0 @@
----
-name: cli-generator
-description: Meta-skill enabling the Agent to automatically convert local software, repositories, or APIs into structured CLI tools using HKUDS/CLI-Anything.
----
-
-# CLI-Generator (Meta-Skill)
-
-> 🔴 **CRITICAL**: This is a Meta-Skill. It gives you (the AI Agent) the ability to forge your own tools to solve complex constraints.
-
-## When to Use
-
-Use this skill when:
-- The user asks you to interact with a complex local GUI software (e.g., GIMP, OBS).
-- You are working with a messy, undocumented local API or scripting repository that is too brittle to run manually via ad-hoc scripts.
-- You need a dedicated CLI tool to perform repetitive tasks reliably, but none exists.
-
-## The Forging Process
-
-When you decide you need a new CLI tool for a given `TARGET_DIR` (the source code of the application), follow exactly these steps:
-
-### 1. Install CLI-Anything (If not already installed)
-Check if `cli-anything` is available in your PATH. If not, install it globally:
-```bash
-pip install cli-anything
-```
-*(If the package is not on PyPI, clone `https://github.com/HKUDS/CLI-Anything` and run `pip install -e .`)*
-
-### 2. Run the Generator Pipeline
-Navigate to a neutral workspace or the target, and run the automated generation pipeline:
-```bash
-cli-anything --source /absolute/path/to/TARGET_DIR
-```
-*Note: This process analyzes the source code and generates the harness. It may take several minutes and uses the underlying LLM's API tokens.*
-
-### 3. Integrate the Forged Skill into `ag-kit`
-The pipeline will generate a new Python package (e.g., `cli-anything-targetname`) and automatically produce a `SKILL.md` file within its package directory (usually `cli_anything/<software>/skills/SKILL.md`).
-You MUST locate this generated `SKILL.md` and copy it into the user's `ag-kit` workspace:
-```bash
-mkdir -p .agent/skills/<targetname-cli>
-cp /path/to/generated/SKILL.md .agent/skills/<targetname-cli>/SKILL.md
-```
-
-### 4. Self-Update
-Once the new `SKILL.md` is in place, you (or any other agent using this repository like GravityClaw / Openclaw) will instantly possess the knowledge to use the newly forged CLI tool. Proceed to solve the user's original objective using the new tool natively.
-
-## Anti-Patterns
-- ❌ **Manual wrappers**: Do NOT try to manually write a CLI harness or wrappers if `cli-anything` can automate it perfectly.
-- ❌ **Token waste**: Do NOT run the generator repeatedly on the same source code. Generate the CLI once, store the `SKILL.md`, and reuse the tool.