菜单
Use when executing implementation plans with independent tasks in the current session
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。
基于 SOC 职业分类
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
Use when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies
Use when you have a written implementation plan to execute in a separate session with review checkpoints
Use when completing tasks, implementing major features, or before merging to verify work meets requirements
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes
Use when starting any conversation - establishes how to find and use skills, requiring Skill tool invocation before ANY response including clarifying questions
| name | subagent-driven-development |
| description | Use when executing implementation plans with independent tasks in the current session |
Execute plan by dispatching fresh browser subagents or breaking work into isolated task boundaries per task, with two-stage review after each: spec compliance review first, then code quality review.
Core principle: Fresh context per task + two-stage review (spec then quality) = high quality, fast iteration
digraph when_to_use {
"Have implementation plan?" [shape=diamond];
"Tasks mostly independent?" [shape=diamond];
"Stay in this session?" [shape=diamond];
"subagent-driven-development" [shape=box];
"executing-plans" [shape=box];
"Manual execution or brainstorm first" [shape=box];
"Have implementation plan?" -> "Tasks mostly independent?" [label="yes"];
"Have implementation plan?" -> "Manual execution or brainstorm first" [label="no"];
"Tasks mostly independent?" -> "Stay in this session?" [label="yes"];
"Tasks mostly independent?" -> "Manual execution or brainstorm first" [label="no - tightly coupled"];
"Stay in this session?" -> "subagent-driven-development" [label="yes"];
"Stay in this session?" -> "executing-plans" [label="no - parallel session"];
}
In Antigravity, you manage the orchestration loop directly:
digraph process {
rankdir=TB;
"Read plan, extract all tasks with full text, note context, create task.md" [shape=box];
"Pick next task" [shape=box];
"Implement task using task_boundary EXECUTION mode" [shape=box];
"Run verification commands" [shape=box];
"Spec compliance self-review: does code match spec?" [shape=diamond];
"Fix spec gaps" [shape=box];
"Code quality self-review: clean, no debt, good names?" [shape=diamond];
"Fix quality issues" [shape=box];
"Mark task complete in task.md" [shape=box];
"More tasks remain?" [shape=diamond];
"Use finishing-a-development-branch skill" [shape=box style=filled fillcolor=lightgreen];
"Read plan, extract all tasks with full text, note context, create task.md" -> "Pick next task";
"Pick next task" -> "Implement task using task_boundary EXECUTION mode";
"Implement task using task_boundary EXECUTION mode" -> "Run verification commands";
"Run verification commands" -> "Spec compliance self-review: does code match spec?";
"Spec compliance self-review: does code match spec?" -> "Fix spec gaps" [label="no"];
"Fix spec gaps" -> "Spec compliance self-review: does code match spec?" [label="re-review"];
"Spec compliance self-review: does code match spec?" -> "Code quality self-review: clean, no debt, good names?" [label="yes"];
"Code quality self-review: clean, no debt, good names?" -> "Fix quality issues" [label="no"];
"Fix quality issues" -> "Code quality self-review: clean, no debt, good names?" [label="re-review"];
"Code quality self-review: clean, no debt, good names?" -> "Mark task complete in task.md" [label="yes"];
"Mark task complete in task.md" -> "More tasks remain?";
"More tasks remain?" -> "Pick next task" [label="yes"];
"More tasks remain?" -> "Use finishing-a-development-branch skill" [label="no"];
}
Since you are the orchestrator in Antigravity (rather than spawning separate agent processes), follow this pattern per task:
The original prompt templates are in this directory:
./implementer-prompt.md — Use as a self-checklist when implementing each task./spec-reviewer-prompt.md — Use as a spec compliance checklist./code-quality-reviewer-prompt.md — Use as a code quality checklistNever:
Always:
Required workflow skills: