sdd-implement

Orchestrate multi-step task implementation with LLM-as-Judge verification.

Updated Mar 10, 2026
One-click install
npx skills add https://github.com/Gamezar/opencode-cek --skill sdd-implement-gamezar
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: sdd-implement
Source: https://github.com/Gamezar/opencode-cek/tree/main/plugins/sdd/skills/sdd-implement
Command: npx skills add https://github.com/Gamezar/opencode-cek --skill sdd-implement-gamezar

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill automates the implementation of complex tasks by orchestrating sub-agents and ensuring quality through LLM-as-Judge verification, reducing manual oversight and improving development speed.

Core Features & Use Cases

  • Orchestrated Task Execution: Manages multi-step task implementation with defined dependencies.
  • Automated Quality Assurance: Integrates LLM-as-Judge for rigorous verification of critical artifacts.
  • Iterative Refinement: Supports --refine mode to re-verify only changed code sections, preserving manual fixes.
  • Use Case: Implement a new feature by providing a task specification file. The Skill will break it down, generate code, and use judges to verify each critical component before completion.

Quick Start

Execute the sdd-implement skill to begin implementing the task defined in 'my-new-feature.task.md'.

Frequently Asked Questions about sdd-implement

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How does LLM-as-Judge verification work for automated task implementation?

Multi-step task implementation is orchestrated by breaking down specifications, generating code via sub-agents, and applying LLM-as-Judge verification to each critical component. It manages task status through a defined folder structure and ensures agent delegation to prevent context overflow.

How do I implement a new feature from a task specification file?

You can refine and re-verify changed code sections iteratively by using the `--refine` mode. This mode specifically re-verifies only the modified code, preserving any manual fixes you have made while ensuring new changes meet quality thresholds.

Can I use automated task orchestration for complex multi-step workflows?

Yes, automated task orchestration manages multi-step task implementation with defined dependencies. It delegates tasks to sub-agents to prevent context overflow and includes human-in-the-loop checkpoints for critical stages of the workflow.

Does iterative refinement preserve manual code fixes during re-verification?

Iterative refinement preserves manual fixes by using the `--refine` mode to re-verify only changed code sections. This ensures that your manual alterations remain intact while the automated LLM-as-Judge verification evaluates the newly modified components.

What are the limitations of using LLM-as-Judge for code verification?

LLM-as-Judge verification relies on configurable quality thresholds and may require human-in-the-loop checkpoints for highly complex logic. While it automates artifact evaluation, users must define appropriate task specifications and folder structures to prevent context overflow during orchestration.