How does Deep Research differ from a Spike?

A Spike is a time-boxed investigation (typically 2-4 hours) into a single technical question or feasibility test. Deep Research is a comprehensive, multi-day process designed for complex analysis that informs major architectural or budgetary decisions.

Does it support task management integration?

Yes, it is designed to work with the ohno plugin to create research tasks, track sub-task progress, and manage follow-up implementation items.

What types of reports does this skill generate?

It produces structured Markdown reports including executive summaries, background context, methodology, comparison matrices, and actionable recommendations with risk assessments.

How is the research output stored?

All research artifacts are organized within your project directory under .claude/research/, including the final report, source logs, and raw notes for full transparency.

Can I use this for vendor evaluations?

Yes, it includes specialized patterns for evaluating SaaS products and tools against weighted criteria like developer experience, cost scalability, and enterprise feature support.

Deep Research

Name: Deep Research
Author: srstomp

bysrstomp

•

生产力与工作流

Conducts comprehensive technical investigations and produces structured reports for high-stakes architectural and business decisions.

The deep-research skill enables Claude to perform multi-day investigations into complex technical topics, framework comparisons, and vendor evaluations. Unlike short-lived spikes, it follows a rigorous five-phase workflow—Scope, Gather, Analyze, Synthesize, and Present—to generate detailed reports that inform major engineering and business stakeholders. It integrates seamlessly with task management tools like ohno to track progress, log findings, and document trade-offs, ensuring that critical decisions are backed by thorough evidence, competitive analysis, and structured rationale.

主要功能

01Structured five-phase research workflow (Scope, Gather, Analyze, Synthesize, Present)

022 GitHub stars

03Standardized templates for comparison matrices and executive summaries

04Tailored workflows for technology, architecture, and vendor evaluations

05Integration with task management for multi-day progress tracking

06Evidence-based analysis focusing on technical trade-offs and migration risks

使用场景

01Performing detailed vendor evaluations for SaaS or infrastructure selection based on weighted criteria

02Conducting competitive analysis and synthesizing industry best practices for system design

03Evaluating and comparing multiple software frameworks or third-party libraries for a new project

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add srstomp/pokayokay deep-research

For use in Claude.ai and ChatGPT