How does it handle verification?

It runs automated success criteria (like test suites) after every phase and pauses for manual human verification before proceeding to the next part of the plan.

Can I use this for small bug fixes?

Yes. The 'Direct Implementation' mode is designed specifically for plans with three or fewer tasks, providing a faster workflow for focused, small-scale changes.

What happens if the codebase doesn't match the plan?

The skill includes a mismatch detection logic that pauses execution, analyzes why the plan cannot be followed, and presents a clear explanation to the user for guidance.

How does Agent Orchestration prevent context loss?

It breaks large plans into individual tasks assigned to sub-agents. Each agent creates a persistent handoff file on disk, allowing the orchestrator to resume the workflow even if the main conversation context is compacted.

Technical Plan Executor

Name: Technical Plan Executor
Author: tfunk1030

bytfunk1030

•

生产力与工作流

Executes and orchestrates complex technical implementation plans with automated verification and context preservation.

The Technical Plan Executor skill transforms static technical plans into actionable workflows by systematically processing roadmaps stored in your repository. It offers two distinct modes: Direct Implementation for quick tasks and Agent Orchestration for large-scale projects. By utilizing a persistent handoff system, the skill ensures that long-running implementations remain resilient even if the AI's context window resets. It enforces rigorous quality control through automated test suites and mandatory human-in-the-loop verification checkpoints, making it ideal for high-stakes feature development and complex refactoring.

主要功能

01Persistent handoff architecture to survive AI context window compaction

02Resumable agent support for clarifying or updating plans mid-execution

03Automatic synchronization with continuity ledgers and markdown plan files

041 GitHub stars

05Dual execution modes for direct implementation or multi-agent orchestration

06Integrated verification loops with automated 'make' checks and manual pauses

使用场景

01Orchestrating complex refactors that require specialized sub-agents per task

02Executing multi-phase feature additions without losing implementation context

03Maintaining a verifiable audit trail of project progress through handoff documentation

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add tfunk1030/vibe implement_plan

For use in Claude.ai and ChatGPT

Download Skill