subagent-driven-development

Name: subagent-driven-development
Availability: InStock
Author: pantheon-org

Dispatch fresh subagents for each task with two-stage reviews.

1|Updated Dec 5, 2025

Search Tags:#workflow automation #plan execution #task orchestration #two-stage review #subagent-driven-development

Ready to run

npx skills add https://github.com/pantheon-org/opencode-plugins --skill subagent-driven-development-pantheon-org

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/pantheon-org/opencode-plugins/tree/main/.opencode/skills/subagent-driven-development
Command: npx skills add https://github.com/pantheon-org/opencode-plugins --skill subagent-driven-development-pantheon-org

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Executing implementation plans can be error-prone and slow when coordinating multiple subagents. This skill solves that by dispatching a fresh subagent for each task and enforcing a two-stage review process to maintain quality and speed.

Core Features & Use Cases

Fresh subagent per task to isolate context and avoid cross-task leakage.
Two-stage review: spec compliance followed by code quality review to ensure correct and high-quality implementations.
Plan-driven execution within a single session, with automatic task extraction and per-task progress tracking.
Suitable for complex, multi-task implementations where tasks are largely independent or require staged validation.

Quick Start

Prepare your implementation plan (e.g., plans/feature-plan.md) and run the Subagent-Driven Development skill to start per-task subagent execution. The system will dispatch a fresh implementer subagent for each task, followed by spec and quality reviews, updating the TodoWrite when tasks are completed.

Frequently Asked Questions about subagent-driven-development

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

FAQPage Schema

How do I automate plan execution with subagents for complex development tasks?▼

Automating plan execution with subagents involves dispatching a fresh subagent for every task within a single session, applying a two-stage review process to maintain quality and speed while isolating context to avoid cross-task leakage.

What is the two-stage review process for subagent task orchestration?▼

The two-stage review process for subagent task orchestration enforces spec compliance followed by code quality review after each task, ensuring correct and high-quality implementations before finalizing development work.

How do I prevent context leakage when orchestrating multiple subagents for implementation plans?▼

Preventing context leakage when orchestrating multiple subagents requires dispatching a fresh subagent for each task, isolating the context window so tasks remain independent and cross-task data does not corrupt execution.

Can I run multi-task implementations in a single session using plan-driven execution?▼

Running multi-task implementations in a single session is supported by plan-driven execution, which automatically extracts tasks, tracks per-task progress, and coordinates reporting without requiring separate sessions.

What is the best way to prepare an implementation plan for subagent-driven development?▼

Preparing an implementation plan for subagent-driven development requires creating a predefined plan file like feature-plan.md, which the system reads to extract tasks, dispatch implementer subagents, and update progress tracking.

When should I not use subagent-driven development for task orchestration?▼

Subagent-driven development for task orchestration should not be used when implementation tasks are highly interdependent or require shared context, as dispatching fresh subagents isolates context and limits cross-task data sharing.