ftm-executor

Execute plan documents through parallel agent worktrees and commit-review-fix loops.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/kkudumu/feed-the-machine --skill ftm-executor
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ftm-executor
Source: https://github.com/kkudumu/feed-the-machine/tree/main/ftm-executor
Command: npx skills add https://github.com/kkudumu/feed-the-machine --skill ftm-executor

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill provides an autonomous end-to-end plan execution engine that coordinates plan interpretation, agent team assembly, and end-to-end plan execution with iterative review loops, enabling large tasks to be completed with minimal user intervention.

Core Features & Use Cases

  • Dynamic agent team creation tailored to plan tasks
  • Parallel worktrees for concurrent task execution with isolation
  • Commit-review-fix loop until all tasks pass verification
  • Use cases: when you need to "execute this plan", "run this plan doc", or "take this plan and implement it"

Quick Start

Provide a plan path to the skill and say go to start autonomous execution.

Frequently Asked Questions about ftm-executor

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

FAQPage Schema
How do I autonomously execute a development plan document?

To execute a plan document autonomously, you provide the plan path to the skill and say go. The engine then analyzes the document, builds a dynamic agent team, and dispatches them in parallel git worktrees to implement tasks.

Can AI agents work on multiple git worktrees in parallel?

Yes, AI agents can work in parallel git worktrees. The engine dispatches assembled agent teams into isolated worktrees, enabling concurrent task execution without conflicts before entering a commit-review-fix loop.

What is a dynamic AI agent team for workflow automation?

A dynamic AI agent team is a group of AI agents automatically assembled and tailored to the specific tasks within a plan document. They execute workflow automation by processing tasks concurrently and iterating through commit-review-fix loops.

How does the commit-review-fix loop work for plan execution?

The commit-review-fix loop iterates through task verification until all plan tasks pass. Agents commit their work in isolated worktrees, undergo automated review, and apply fixes until the entire plan execution is complete.

Do I need to manually intervene during autonomous plan execution?

No, autonomous plan execution requires minimal user intervention. Once you provide the plan document path and start the process, the engine handles agent team assembly, parallel task execution, and iterative review loops independently.