What happens if the decomposition script fails to find tasks?

Ensure your plan uses clear markers like '## Task 1:' or '## 1.'. You can also run the script with a verbose flag to debug the parsing logic or manually add task markers.

What is the benefit of breaking plans into separate task files?

Individual task files contain only the context relevant to that specific step, reducing the tokens a subagent must process by up to 90%, resulting in higher accuracy and lower costs.

How are dependencies managed during decomposition?

The skill analyzes your plan for explicit dependency mentions and automatically detects file-based conflicts where multiple tasks modify the same file, ensuring a safe execution order.

How does this skill improve development speed?

It identifies tasks that can run simultaneously without conflict, allowing up to two subagents to work in parallel, which can reduce total implementation time by up to 40%.

Plan Decomposition & Parallelization

Name: Plan Decomposition & Parallelization
Author: seanGSISG

byseanGSISG

生産性とワークフロー

Breaks down monolithic implementation plans into atomic, parallelizable tasks for efficient AI subagent execution.

概要

This skill streamlines the transition from high-level planning to granular execution by automatically splitting monolithic markdown plans into discrete task files and a machine-readable execution manifest. By intelligently analyzing file-based dependencies and explicit task relationships, it identifies opportunities to run up to two subagents simultaneously. This process significantly reduces context window bloat—often by up to 90% per task—while accelerating development timelines through parallelized workflows.

主な機能

Automated parsing of monolithic plan files into discrete task markdown files
0 GitHub stars
Intelligent dependency analysis based on file conflicts and explicit task markers
Generation of execution manifests (JSON) for parallel batch scheduling
Significant context token reduction for subagent-driven development
Built-in verification checklists for every decomposed task

ユースケース

Transforming complex feature roadmaps into actionable, structured development tasks
Optimizing multi-file refactorings for parallel processing by AI subagents
Reducing LLM context usage during large-scale code implementation phases

概要

主な機能

Automated parsing of monolithic plan files into discrete task markdown files
0 GitHub stars
Intelligent dependency analysis based on file conflicts and explicit task markers
Generation of execution manifests (JSON) for parallel batch scheduling
Significant context token reduction for subagent-driven development
Built-in verification checklists for every decomposed task

ユースケース

Transforming complex feature roadmaps into actionable, structured development tasks
Optimizing multi-file refactorings for parallel processing by AI subagents
Reducing LLM context usage during large-scale code implementation phases