AI Engineering

Coding Agent Standards defines practical standards for implementation-focused coding agents. Use it when creating or editing code, especially when reliability, clarity, and low-risk delivery are required.

Component Interfaces discovers, parses, and loads coding skills from SKILL.md files, controlling visibility and enabling policies. Use it to integrate skill metadata and prompt fragments into an agent harness.

Prompt Architect turns vague objectives into clear, testable prompts for agent workflows. Use it when creating new prompts, refining weak prompts, or establishing reusable prompting patterns.

Aider Delegate hands bounded coding tasks to Aider as a background implementer and reviews its diff before landing. Use this to delegate implementation, fixes, or refactors to Aider or a local/self-hosted OpenAI-compatible endpoint.

Antigravity Delegate hands a bounded coding task to the Google Antigravity CLI (`agy`) as a background implementer, then lets the orchestrator review the diff and commit the verified work. Use it when you want to delegate implementation, fixes, or refactors to Antigravity and stay the reviewer, or run a queue of coding tasks through agy.

Cline Delegate delegates a bounded coding task to the Cline coding agent CLI as a background implementer, then has the orchestrator review the diff and commit it. Use it when the user wants to delegate implementation work to Cline or run a queue of coding tasks while staying the reviewer.

Codex Delegate hands a bounded coding task to the OpenAI Codex CLI to implement in the background, then lets you review the diff and commit it yourself. Use it when the user wants to delegate implementation work to Codex or run a queue of coding tasks while staying the reviewer.

Copilot Delegate hands a bounded coding task to the GitHub Copilot CLI as a background implementer, then reviews the diff and commits it. Use when the user wants to delegate implementation work to Copilot or run a queue of coding tasks while staying the reviewer.

Cursor Delegate dispatches coding tasks to the Cursor Agent CLI for implementation, then reviews the diff and commits the result. Use it when you want to hand off coding work to Cursor and act as the reviewer.

Delegate Setup discovers installed implementer CLIs and writes delegation lane configuration after user approval. Use it to define which CLI handles feature, test, or UI work, with optional model and effort dials.

Grok Delegate hands a bounded coding task to the Grok Build CLI as a background implementer, then has the orchestrating agent review the diff and commit it. Use it when delegating implementation, fixes, or refactors to Grok and reviewing the result yourself.

Kimi Delegate hands a bounded coding task to the Kimi Code CLI as a background implementer, then reviews its diff and commits the result. Use it when the user wants to delegate implementation work to Kimi or run a queue of coding tasks through Kimi while staying the reviewer.