skills.rest
caseyrtalbot
caseyrtalbotCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

drive-execution

Dispatch isolated implementers to execute plan tasks with per-task verification.

Updated Aug 27, 2026
Search Tags:#automation#verification#subagents#orchestrate#task-dispatch#execution-workflow#prove-first
One-click install
npx skills add https://github.com/caseyrtalbot/claude-plugins --skill drive-execution
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: drive-execution
Source: https://github.com/caseyrtalbot/claude-plugins/tree/main/plugins/forge/skills/drive-execution
Command: npx skills add https://github.com/caseyrtalbot/claude-plugins --skill drive-execution

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Orchestrates end-to-end plan execution by dispatching dedicated implementers for each task, ensuring per-task verification and preventing cross-task contamination of context.

Core Features & Use Cases

  • Per-task delegation: spawn a fresh implementer for each plan task to enforce isolation.
  • Prove-first workflow: require tests before acceptance to ensure quality.
  • End-to-end verification: engage a quality auditor before marking tasks complete.

Quick Start

Dispatch a complete plan, spawn per-task implementers, and run the quality audit before confirming completion.

Frequently Asked Questions about drive-execution

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

FAQPage Schema
How do I orchestrate task-by-task execution with isolated agents?▼

Task-by-task verification engages a quality auditor to review each task before marking it complete. This auditor checks the work generated by the isolated implementer agent to ensure it meets the plan's requirements.

What is a prove-first workflow for software execution?▼

You need a complete and ordered plan where tasks can be run by isolated agents. The workflow dispatches per-task implementers to execute this existing plan, requiring a fully structured task list before execution begins.

Why does cross-task context contamination happen during plan execution?▼

After per-task implementers complete their assigned tasks and a quality auditor verifies the work, the workflow performs a final confirm-complete transition. This final step officially marks the entire end-to-end plan execution as finished.

Can I use isolated subagents to execute an existing ordered plan?▼

Yes, you can use isolated subagents to execute an existing ordered plan by dispatching a fresh implementer for each task. This task dispatch approach ensures independent verification and prevents cross-task contamination throughout the execution workflow.

Related Skills

Explore More Software Engineering Skills→
gqf2008gqf2008

subagent-driven-development

Dispatch tasks to fresh subagents with isolated context and two-stage reviews.

Community
Advanced
hffmnnjhffmnnj
39

security-audit

Audit code for security vulnerabilities and generate a structured security report.

Community
Advanced
greyhaven-aigreyhaven-ai
29

grey-haven-tdd-orchestration

Coordinate multi-agent TDD workflows enforcing red-green-refactor discipline across Python and TypeScript.

Community
Advanced
Veloxia-agencyVeloxia-agency

spawn

Coordinate parallel subagents in isolated git worktrees for shared tasks.

Community
Advanced