skills.rest
nuts-and-bolts-ai
nuts-and-bolts-aiOFFICIAL
Explore all skills by this author →
GitHub

subagent-driven-development

Dispatch independent subagents to execute implementation plans with two-stage reviews.

Updated Mar 8, 2026
Search Tags:#development#code review#code generation#ai agents#implementation plan#task execution
One-click install
npx skills add https://github.com/nuts-and-bolts-ai/nanoclaw-cheerful --skill subagent-driven-development-nuts-and-bolts-ai
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/nuts-and-bolts-ai/nanoclaw-cheerful/tree/main/groups/slack_cheerful-dev/.claude/skills/subagent-driven-development
Command: npx skills add https://github.com/nuts-and-bolts-ai/nanoclaw-cheerful --skill subagent-driven-development-nuts-and-bolts-ai

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill automates the process of executing implementation plans by breaking them down into independent tasks, each handled by a dedicated AI subagent, ensuring high-quality code through a structured, two-stage review process.

Core Features & Use Cases

  • Task Decomposition: Breaks down complex plans into manageable, independent tasks.
  • Subagent Execution: Dispatches a fresh AI subagent for each task to prevent context pollution.
  • Two-Stage Review: Implements a rigorous review process: first for spec compliance, then for code quality.
  • Iterative Refinement: Manages feedback loops for implementer subagents to fix issues identified during reviews.
  • Use Case: When you have a detailed plan to add a new feature to your codebase, this Skill can autonomously execute each step, from writing the code to ensuring it meets specifications and quality standards, significantly speeding up development cycles.

Quick Start

Use the subagent-driven-development skill to execute the implementation plan found in the file 'docs/superpowers/plans/feature-plan.md'.

Frequently Asked Questions about subagent-driven-development

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

FAQPage Schema
How do I execute an implementation plan using AI agents?▼

Executing an implementation plan with AI agents involves breaking down complex plans into independent tasks, dispatching a fresh subagent for each to prevent context pollution, and applying a two-stage review for spec compliance and code quality.

What is subagent-driven development and how does it prevent context pollution?▼

Subagent-driven development is a technique that dispatches a fresh AI subagent for each independent task within an implementation plan, isolating context to prevent pollution and managing task statuses like DONE, BLOCKED, and NEEDS_CONTEXT with defined escalation paths.

How do AI agents handle code review and iterative refinement during task execution?▼

AI agents handle code review through a two-stage process checking spec compliance then code quality, managing iterative refinement by routing feedback loops back to implementer subagents to fix identified issues before finalizing task statuses.

Can I optimize model selection for AI code generation based on task complexity?▼

Yes, you can optimize model selection for AI code generation based on task complexity. The system manages model selection to optimize for cost and speed efficiency, matching resources to the specific demands of each independent subagent task.

What are the limitations of using autonomous AI agents for code generation?▼

Limitations include encountering task statuses like BLOCKED or NEEDS_CONTEXT, which require defined escalation paths and human intervention when subagents cannot resolve spec compliance or code quality issues through automated iterative refinement loops.

Related Skills

Explore More Software Engineering Skills→
GregoryHoGregoryHo
6

arc-requesting-review

Dispatch a code-reviewer subagent with implementation details and commit SHAs.

Community
Intermediate
tunneleventunneleven

subagent-driven-development

Orchestrate software development tasks with subagents for coding and review.

Community
Advanced
ToqsickToqsick

subagent-driven-development

Orchestrate subagent execution with a two-stage review for spec compliance and code quality.

Community
Advanced
roderikroderik
2

subagent-driven-development

Dispatch fresh subagents per task to execute implementation plans with two-stage reviews.

Community
Advanced