What is the Research Workflow Management skill?

It is a specialized capability for Claude Code that provides a structured framework for software development, guiding you through researching, planning, experimenting, implementing, and validating changes.

Where are the workflow documents stored?

All research notes, implementation plans, and validation reports are saved as structured markdown files in the .agents/ directory of your project's root folder.

Can I skip certain steps in the workflow?

Yes, the workflow is designed to be flexible. You can skip optional phases like Experimentation for straightforward tasks or use Research-only mode for codebase exploration.

How does this skill help with code validation?

The Validate phase systematically compares the final implementation against the specific success criteria defined during the Planning phase, generating a report of pass/fail statuses.

Is this workflow compatible with Git?

Yes, the generated markdown documents provide an excellent audit trail that can be committed to your repository to document the 'why' and 'how' of your technical decisions.

Research Workflow Management

Name: Research Workflow Management
Author: uw-ssec

byuw-ssec

•

生产力与工作流

Orchestrates a structured, six-phase development lifecycle from initial codebase exploration to final implementation validation.

This skill provides a systematic, AI-enabled framework for managing complex software development tasks by breaking them into distinct, auditable phases: Research, Plan, Iterate, Experiment, Implement, and Validate. By generating structured documentation in a local .agents/ directory, it ensures that every technical decision is backed by evidence and every implementation follows a verified roadmap. This methodology reduces cognitive load for developers and improves code quality by separating architectural discovery from execution, creating a navigable graph of technical decisions and their eventual implementations.

主要功能

01Six-phase structured workflow including Research, Planning, and Validation

02Interactive planning tools with measurable automated and manual success criteria

03Systematic validation reports to verify final code against original plan requirements

04Automated generation of technical documentation and implementation roadmaps

05Experimental prototyping phase for evidence-based architectural decision making

0618 GitHub stars

使用场景

01Mapping and documenting complex legacy codebases before starting major refactoring projects

02Designing and implementing new features with clear success criteria and auditable decision trails

03Conducting comparative technical experiments to determine the best approach for architectural changes

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add uw-ssec/rse-plugins research-workflow-management

For use in Claude.ai and ChatGPT

Download Skill