gsd-execute-phase

Execute phase plans in parallel using wave-based subagent orchestration.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/optivent/gsd-kimi-cli --skill gsd-execute-phase
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gsd-execute-phase
Source: https://github.com/optivent/gsd-kimi-cli/tree/main/skills/gsd-execute-phase
Command: npx skills add https://github.com/optivent/gsd-kimi-cli --skill gsd-execute-phase

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill orchestrates the execution of multiple phase plans in parallel using a wave-based strategy, reducing overall cycle time and improving reliability by concurrent task handling and context sharing.

Core Features & Use Cases

  • Wave-based parallel execution of multiple PLAN.md files within a phase.
  • Deterministic task handling via subagents (gsd-executor) ensuring atomic plan execution.
  • Context-aware progress and recovery with STATE.md and plan-specific SUMMARY.md generation.

Quick Start

Run a phase with a single instruction: /skill:gsd-execute-phase <phase-number>.

Frequently Asked Questions about gsd-execute-phase

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

FAQPage Schema
How do I execute multiple phase plans in parallel to reduce cycle time?

Parallel execution of phase plans is orchestrated using a wave-based approach, grouping tasks from multiple PLAN.md files into concurrent waves to reduce overall cycle time and improve reliability.

What is wave-based parallel execution for multi-phase project workflows?

Wave-based parallel execution is a strategy that groups tasks from multiple phase plans into sequential waves, allowing concurrent execution within each wave while maintaining deterministic task handling and atomic plan execution.

How do I run parallel execution for a specific phase using a single command?

Run phase execution by invoking the skill with a phase number, which triggers automated plan discovery, wave grouping, parallel executor subagents, and progress reporting across all plans within that phase.

Do I need specific configuration files to set up parallel phase plan orchestration?

Yes, parallel phase orchestration requires a .planning/config.json configuration file and individual PLAN.md files for each phase to properly handle plan discovery and wave-based execution.

How does context-aware progress tracking work during parallel phase execution?

Context-aware progress tracking generates STATE.md updates and plan-specific SUMMARY.md files during parallel execution, enabling progress reporting and recovery across concurrent phase workflows.

What happens if a task fails during wave-based parallel execution of phase plans?

Wave-based execution handles task failures through context-aware recovery using STATE.md and SUMMARY.md files, allowing deterministic subagents to maintain atomic plan execution and track progress across phases.