gsd-discuss-phase

Gather implementation decision context by adaptively questioning users before planning.

Updated Feb 11, 2026
One-click install
npx skills add https://github.com/ArtRiv/game-topicos-especiais --skill gsd-discuss-phase-artriv
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: gsd-discuss-phase
Source: https://github.com/ArtRiv/game-topicos-especiais/tree/main/.github/skills/gsd-discuss-phase
Command: npx skills add https://github.com/ArtRiv/game-topicos-especiais --skill gsd-discuss-phase-artriv

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It gathers implementation-critical context by adaptively questioning the user so downstream agents can proceed without rediscovering or revisiting already-decided choices.

Core Features & Use Cases

  • Implementation decision capture: Distills “gray areas” into explicit decisions that researcher and planner can rely on.
  • Context-aware execution: Loads prior phase artifacts (e.g., PROJECT.md, REQUIREMENTS.md, STATE.md, and prior CONTEXT.md files) to avoid re-asking settled questions.
  • Interactive or non-interactive modes: Supports routing between a assumptions-first flow and a full discussion flow, including an auto mode to skip interactive selection.
  • Decision-focused output: Produces a phase-specific {phase_num}-CONTEXT.md designed to guide subsequent research and planning.

Quick Start

Run the gsd-discuss-phase skill with a required phase number in non-interactive mode by using the --auto flag to generate the {phase_num}-CONTEXT.md output.

Frequently Asked Questions about gsd-discuss-phase

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

FAQPage Schema
How do I capture implementation decisions before planning a new software phase?▼

Capturing implementation decisions requires adaptively questioning the team to resolve gray areas before planning. This skill loads prior phase artifacts like PROJECT.md and STATE.md, scouts the codebase, and asks targeted questions to distill explicit choices into a phase-specific CONTEXT.md for downstream agents.

What is adaptive questioning for codebase context building?▼

Adaptive questioning for codebase context building dynamically identifies unresolved gray areas by analyzing prior artifacts and codebase state. It routes between interactive discussion and assumptions-first flows, ensuring only unsettled decisions are queried rather than re-asking established requirements.

How do I avoid re-asking settled questions when preparing project context?▼

To avoid re-asking settled questions, load prior phase artifacts including PROJECT.md, REQUIREMENTS.md, and previous CONTEXT.md files. This context-aware execution scouts the codebase to identify only remaining gray areas, skipping previously resolved decisions and generating a focused phase-specific context file.

Can I generate a CONTEXT.md file automatically without interactive selection?▼

Yes, you can generate a CONTEXT.md file automatically by running the skill in non-interactive mode using the --auto flag. This bypasses interactive gray area selection and produces a phase-specific context output based on existing artifacts and codebase scouting.

Does workflow routing support different discussion modes for decision capture?▼

Workflow routing supports multiple decision capture modes including an interactive full discussion flow, an assumptions-first flow, and an auto mode to skip interactivity. Routing behavior is configured via config files, determining how gray areas are processed and resolved for the output context.

What are the limitations of using automated decision capture for planning preparation?▼

Automated decision capture relies entirely on existing prior artifacts and codebase scouting quality. If PROJECT.md, REQUIREMENTS.md, or STATE.md files are incomplete or outdated, the adaptive questioning may misidentify gray areas, and the auto mode will generate assumptions without user validation for downstream agents.