skills.rest
Junkfooooood
JunkfoooooodCOMMUNITY
Explore all skills by this author →
GitHub

subagent-driven-development

Executes implementation plans by dispatching fresh subagents per task with two-stage review.

Updated May 4, 2026
Search Tags:#workflow#tdd#code review#implementation#delegation#plan execution#subagent
One-click install
npx skills add https://github.com/Junkfooooood/hermes-governance --skill subagent-driven-development-junkfooooood
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/Junkfooooood/hermes-governance/tree/main/skills/software-development/subagent-driven-development
Command: npx skills add https://github.com/Junkfooooood/hermes-governance --skill subagent-driven-development-junkfooooood

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Executing a multi-task implementation plan in a single session leads to context pollution, inconsistent quality, and undetected spec drift. This Skill orchestrates plan execution through isolated subagents with enforced review gates so every task is verified before moving on. ## Core Features & Use Cases - Per-Task Subagent Dispatch: Each plan task runs in a fresh subagent with full task text and project context provided directly, avoiding accumulated state confusion. - Two-Stage Review Gates: Every task passes a spec compliance review first, then a code quality review, with fix-and-re-review loops until approved. - Final Integration Review: After all tasks complete, a reviewer checks cross-task consistency, runs the full test suite, and confirms merge readiness. - Use Case: You have a 5-task authentication feature plan. The Skill dispatches an implementer subagent for the User model, verifies it against the spec, reviews code quality, then repeats for password hashing, login endpoint, and remaining tasks before a final integration check. ## Quick Start Execute the implementation plan in docs/plans/feature-plan.md using subagent-driven development with spec and quality reviews for each task.

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 with subagents?▼

Read the plan once, extract all tasks into a todo list, then dispatch a fresh implementer subagent per task with the full task text in context. Follow each implementation with a spec compliance review and a code quality review before marking the task complete.

What is the correct order for spec review and code quality review?▼

Spec compliance review always runs first, followed by code quality review. Starting quality review before spec compliance passes is explicitly prohibited because fixing spec gaps may invalidate quality findings.

Why use a fresh subagent for each task instead of one session?▼

Fresh subagents prevent context pollution from accumulated state, giving each task clean, focused context without confusion from prior tasks' code or reasoning. The trade-off is more subagent invocations, but issues are caught earlier.

What happens when a reviewer finds issues in a task?▼

The implementer subagent (or a new fix subagent) addresses the issues, then the reviewer re-reviews. This loop repeats until approval, and you never proceed to the next task with open critical or important issues.

When should I not use subagent-driven development?▼

Avoid it when you have no implementation plan, when tasks are tightly coupled and touch the same files concurrently, or when you would skip the review gates. Dispatching parallel implementers on shared files is a listed red flag.

Related Skills

Explore More Software Engineering Skills→
aist-ai-sastaist-ai-sast

aist-plan

Decompose complex development work into atomic 2-5 minute test-first tasks.

Official
Intermediate
blakesimsblakesims

review-merge

Validate feature branch merge readiness with integration checks and summaries.

Community
Advanced
tiankong0101-bytetiankong0101-byte

superpowers

Enforce a spec-first workflow with brainstorming, TDD, debugging, and branch completion.

Community
Advanced
databrickslabsdatabrickslabs
38

subagent-driven-development

Execute multi-task implementation plans with fresh subagents and two-stage reviews.

Official
Advanced