gsd-research-phase

Validate a project phase and spawn a subagent to produce RESEARCH.md.

Updated Jan 25, 2026
One-click install
npx skills add https://github.com/MINT-IA/MINT --skill gsd-research-phase-mint-ia
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gsd-research-phase
Source: https://github.com/MINT-IA/MINT/tree/main/.claude/skills/gsd-research-phase
Command: npx skills add https://github.com/MINT-IA/MINT --skill gsd-research-phase-mint-ia

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill automates the technical research step for a project phase by validating the phase, collecting relevant phase context, checking prior research, and launching a dedicated subagent to produce a focused RESEARCH.md, so engineering teams can separate investigation from planning and preserve main-context clarity.

Core Features & Use Cases

  • Phase validation and discovery: Confirms the requested phase exists on the roadmap and extracts phase metadata before any work begins.
  • Existing research handling: Detects prior RESEARCH.md and offers update, view, or skip options to avoid duplicated effort.
  • Subagent orchestration: Spawns a gsd-phase-researcher with resolved model settings, provides file paths for reading, enforces quality gates, and writes structured research output for downstream planning.
  • Use Case: Run a targeted feasibility or ecosystem study for Phase 3 to collect architecture patterns, standard stacks, pitfalls, and code examples without bloating the primary conversation context.

Quick Start

Run gsd-research-phase with the phase identifier to validate the phase, gather context, and spawn a gsd-phase-researcher that outputs a phased RESEARCH.md.

Frequently Asked Questions about gsd-research-phase

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

FAQPage Schema
How do I automate technical research for a specific project phase without bloating the main context?

This Skill isolates technical investigation by spawning a focused research subagent that validates the phase against the roadmap, gathers requirements, and outputs a phase RESEARCH.md, preserving main-context clarity for the engineering team.

What is the best way to run a feasibility check for a project phase before integration?

The best way to run a feasibility check is to validate the phase against your roadmap, detect any prior RESEARCH.md to avoid duplicated effort, and spawn a dedicated subagent to perform an isolated technical investigation before planning.

How does subagent orchestration work for isolated technical investigation?

Subagent orchestration works by spawning a phase researcher with resolved model settings, passing relevant file paths for context reading, enforcing quality gates, and writing structured research output for downstream planning integration.

Can I update an existing RESEARCH.md file instead of starting a new technical investigation?

Yes, when prior research is detected, the system offers update, view, or skip options to avoid duplicated effort, allowing you to either append new findings to the existing RESEARCH.md or re-research from scratch.

Do I need a roadmap defined before running phase research?

Yes, a roadmap is required. Phase validation confirms the requested phase exists on the roadmap and extracts phase metadata before any research begins, ensuring the investigation aligns with established project planning workflows.

When should I separate technical research from project planning workflows?

You should separate technical research from planning when a phase requires isolated investigation, re-research, or feasibility checks before integrating findings into phase plans, preventing context bloat and preserving primary conversation clarity.