Does Plan Coordinator execute the code itself?

No, Plan Coordinator is a planning-only agent. It builds the execution roadmap and assigns agents, but the actual delegation and execution are handled by the main Claude context.

How does it decide which agent to assign to a task?

It parses the YAML frontmatter of your available agents in the .claude/agents/ directory and matches their stated expertise to the specific requirements of each plan step.

Can it handle parallel execution?

Yes, it identifies steps with shared dependencies that can run simultaneously and groups them into parallel waves to speed up the implementation process.

What is a 'Step Bundle' in the Execution Map?

A bundle is a group of sequential, tightly coupled tasks assigned to the same agent that can be completed in a single delegation call to maximize efficiency.

Where does it store the execution progress?

Progress is written to markdown files within your project's docs/plans/ directory, ensuring the execution state is preserved even if the conversation context is reset.

Plan Coordinator

Name: Plan Coordinator
Author: odeciojunior

byodeciojunior

0•

生产力与工作流

Transforms complex implementation plans into structured Execution Maps for multi-agent coordination and systematic task delegation.

The Plan Coordinator is a specialized utility designed to bridge the gap between high-level planning and technical execution. It analyzes implementation plans, maps specific steps to the most qualified executor agents, and organizes tasks into efficient parallel groups or sequential bundles. By generating persistent progress files and detailed context blocks, it ensures that the main Claude instance has a clear, state-aware roadmap for delegation, preventing context loss and maintaining project momentum during complex multi-step implementations.

主要功能

01Smart step bundling based on turn budgets and domain coupling

02Automated parallelization of independent tasks to optimize execution speed

03Rigorous prerequisite verification and dependency resolution before execution

04Persistent state management via progress files for recovery across sessions

050 GitHub stars

06Automatic Execution Map generation with intelligent agent-to-task matching

使用场景

01Managing complex code refactoring projects that require coordinated sequential steps

02Orchestrating large-scale feature implementations across frontend, backend, and testing domains

03Maintaining execution state and recovery points for long-running implementation cycles

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add odeciojunior/claude-play plan-coordinator

For use in Claude.ai and ChatGPT

主要功能

01Smart step bundling based on turn budgets and domain coupling

02Automated parallelization of independent tasks to optimize execution speed

03Rigorous prerequisite verification and dependency resolution before execution

04Persistent state management via progress files for recovery across sessions

050 GitHub stars

06Automatic Execution Map generation with intelligent agent-to-task matching

使用场景

01Managing complex code refactoring projects that require coordinated sequential steps

02Orchestrating large-scale feature implementations across frontend, backend, and testing domains

03Maintaining execution state and recovery points for long-running implementation cycles

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add odeciojunior/claude-play plan-coordinator

For use in Claude.ai and ChatGPT