spec-gen

Generate OpenSpec proposals from TaskMaster subtasks in isolated batches.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/turbobeest/squawk --skill spec-gen
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: spec-gen
Source: https://github.com/turbobeest/squawk/tree/main/.claude/skills/spec-gen
Command: npx skills add https://github.com/turbobeest/squawk --skill spec-gen

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill bridges the gap between high-level tasks and actionable implementation specs by automatically creating OpenSpec proposals for every subtask. It ensures nothing is lost in translation from planning to coding.

Core Features & Use Cases

  • Batch Processing: Efficiently processes tasks in isolated batches to manage context and prevent errors.
  • 1:1 Mapping: Creates one clear, detailed OpenSpec proposal for each subtask, ready for TDD implementation.
  • Sequential Safety: Processes batches sequentially to avoid context exhaustion and ensure reliability.
  • Use Case: After task decomposition, you have 45 subtasks. This Skill processes them in batches of 2 master tasks, creating 45 corresponding OpenSpec proposal files with titles, descriptions, and test strategies pulled from the tasks.

Quick Start

Activate the skill to begin generating OpenSpec proposals from all subtasks in the decomposed tasks.json.

Frequently Asked Questions about spec-gen

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

FAQPage Schema
How do I automatically generate specifications from decomposed tasks?

Specification generation creates one detailed OpenSpec proposal per subtask from your task decomposition. Activate the Skill to process all subtasks in tasks.json, producing proposal files with titles, descriptions, and test strategies ready for implementation.

What's the best way to convert task planning into actionable implementation specs?

Task-to-spec conversion bridges planning and coding by automatically generating OpenSpec proposals for every subtask. This ensures specifications align with decomposed work and prevents information loss between planning and development phases.

Can I process large numbers of subtasks without context exhaustion?

Yes, batch processing handles large task volumes by processing 5–10 master tasks sequentially in isolated batches. Each batch creates proposals in separate worktrees, preventing context exhaustion and ensuring reliable generation across 45+ subtasks.

How does batch isolation improve specification generation reliability?

Batch isolation runs each group of tasks in a separate worktree with atomic commits, preventing cross-batch errors and context interference. Sequential execution ensures consistency while the 1:1 mapping guarantees one proposal per subtask.

What happens to specifications after batch processing completes?

Upon phase transition, the Skill signals completion after updating TASKMASTER_OPENSPEC_MAP.md with all proposal references. Generated specifications are ready for test-driven development and downstream implementation workflows.