Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    samjil delegate agy

    1

    Delegates implementation, testing, and refactoring to Antigravity CLI to save Claude quota.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    samjil delegate agy

    samjil delegate agy

    Example session with this skill installed

    게시판 API의 유효성 검사 로직을 추가해야 해. 설계는 이미 끝났고, Validator 클래스만 구현하면 돼. agy에게 위임해줘. @cwd: D:\Projects\board-api

    • Read your context and instructions
    • Compiled the samjil delegate agy

    (agy(gemini-3.8-flash-medium)에게 위임하고 Claude가 검토한 결과입니다)

    게시판 요청 데이터에 대한 유효성 검사 로직을 구현하였습니다. RequestValidator.ts 파일을 생성하고 제목, 내용, 작성자 필드에 대한 제약 조건을 적용한 뒤 단위 테스트를 통과시켰습니다.

    구현 결과

    // src/validators/RequestValidator.ts
import { ValidationError } from '../errors/AppError';

export interface PostRequest {
  title: string;
  content: string;
  author: string;
}

export const validatePostRequest = (data: PostRequest): void => {
  if (!data.title || data.title.length < 5 || data.title.length > 100) {
    throw new ValidationError('Title must be between 5 and 100 characters.');
  }
  if (!data.content || data.content.length < 10) {
    throw new ValidationError('Content must be at least 10 characters long.');
  }
  if (!data.author || data.author.trim() === '') {
    throw new ValidationError('Author is required.');
  }
};

    Next steps

    • src/controllers/PostController.ts에서 위 Validator를 호출하도록 연결합니다.
    • git diff를 통해 생성된 파일의 위치와 네이밍 컨벤션이 적절한지 최종 확인합니다.
    • 변경 사항이 만족스러우면 git commit을 진행합니다.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Stop wasting your Claude quota on boilerplate, repetitive implementation, and low-level debugging. This skill delegates routine tasks to the Antigravity CLI (agy) watcher, ensuring you only spend high-value tokens on architecture and complex logic. It automates the handoff process while maintaining a strict verification loop to keep quality high.

    What it does

    • Smart Delegation partitions work by difficulty, sending implementation, refactoring, and testing to agy while keeping architecture design for the primary agent.
    • Project Context Management uses @cwd to grant the watcher safe, direct access to your local project directories for file-based operations.
    • Automated Verification enforces a mandatory review process where every delegated output is checked against design specs before acceptance.
    • Quota Optimization calculates token trade-offs before every task to ensure delegation is actually cheaper than direct processing.
    • Session Continuity maintains context across multiple turns using @session tags, allowing complex multi-step workflows.

    How it works

    1. Analyze Task: The skill evaluates every request to determine if it is a "design/high-difficulty" task or a delegatable implementation.
    2. Handoff Brief: If delegatable, it generates a 4-part structured brief including goals, file paths, constraints, and verification commands.
    3. Execution: It triggers submit-task.ps1 to send the brief to the agy watcher for background processing.
    4. Review & Refine: The agent inspects the diff and test results from agy. If it meets the quality bar, it commits; otherwise, it sends feedback for a re-run.

    Frameworks & tools

    Antigravity CLI (agy), PowerShell 런타임, and local file system access via @cwd. Works with any tech stack (Python, React, SQL, etc.) that agy can process.

    Why this beats prompting it yourself

    Manual handoffs are brittle and often lead to lost context or poor-quality code. This skill uses a standardized 4-part brief system and automated "git-guard" snapshots to ensure delegation is safe, predictable, and significantly cheaper than manual prompting.

    Use cases

    • Feature Implementation: Generating UI components or API endpoints based on a pre-defined architecture.
    • Systematic Refactoring: Cleaning up technical debt or applying linting fixes across multiple files.
    • Test Generation: Writing unit and integration tests for existing modules.
    • Documentation: Updating READMEs and docstrings based on code analysis.

    Known limitations

    The agy watcher must be running locally. It lacks web search capabilities and cannot access external authenticated browser sessions or Claude-specific artifacts.

    Requires agy watcher running on Windows/PowerShell. Cannot perform web searches. Restricted to local file operations and CLI tools. Context is limited to provided @cwd and @session history.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed today

    Requires

    macOS
    Git

    Permissions

    Terminal / Shell
    Listedtoday

    What's inside

    Frequently Asked Questions

    The skill automates the handoff of routine coding tasks (like boilerplate and testing) to the Antigravity CLI, allowing you to save your Claude usage quota for high-level architectural decisions and complex logic.

    You need the Antigravity CLI (agy) installed and running locally, a PowerShell runtime, and an agent environment that supports the @cwd and @session tags for file system access and context management.

    While it manages the implementation, it includes a mandatory 'Review & Refine' step where the agent inspects diffs and test results against your original specs before any code is finalized.

    Yes, the skill is language-agnostic; it works with any stack—including Python, React, and SQL—that the Antigravity CLI can process within your local directory.

    You receive the specialized skill configuration, the structured 4-part brief system for task handoffs, and the submit-task.ps1 integration script to connect your agent to the agy watcher.