メニュー
Use when executing implementation plans with independent tasks in the current session. Subagents follow a two-stage review loop.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
| name | subagent-driven-dev |
| description | Use when executing implementation plans with independent tasks in the current session. Subagents follow a two-stage review loop. |
| type | skill |
Execute an implementation plan by dispatching a fresh specialized agent (e.g., @Bug-Hunter or @Operator) per task, with a two-stage review after each: spec compliance review first, then code quality review.
Why subagents: You delegate tasks to specialized agents with isolated context. By precisely crafting their instructions and context, you ensure they stay focused and succeed at their task. They should never inherit your session's history — you construct exactly what they need.
implementation_plan.md and extract all tasks.@Architect (SYARCH) or spec-reviewer subagent to confirm code matches the spec.
code-reviewer-specialist to review for SOLID principles and Antigravity Clean Core alignment.
Implementer subagents report one of four statuses. Handle each appropriately:
Never ignore an escalation.
You: I'm using Subagent-Driven Development to execute this plan.
[Extract all 5 tasks from docs/plans/feature-plan.md]
Task 1: Hook installation script
[Dispatch @Operator subagent with full task text + context]
Implementer: "Added install-hook command. Tests passing."
[Dispatch spec compliance reviewer]
Spec reviewer: ✅ Spec compliant
[Dispatch code-reviewer-specialist]
Code reviewer: ✅ Approved
[Mark Task 1 complete, move to Task 2]
Never:
If reviewer finds issues, the implementer MUST fix them, and the reviewer MUST re-review.
Automated governance, hook installation, pre-commit validation, branch isolation, and safe commit operations.
Enforcement of safety guardrails, axiom verification, secret scanning, and mutability protections.
Building and managing premium statistical dashboards, RAG explorer UIs, warehouse log monitors, and real-time visualization centers.
Integration of agent systems with blockchain protocols, contracts, event stores, reputation networks, and collective verification engines.
AI model configuration, LLM memoization, agentic RAG search, and vector library maintenance.
Catalog generation, reference link verification, workshop documentation building, and knowledge gap analysis.