Can it investigate external sources?

Yes, for complex queries, the workflow can spawn domain-specific agents like a web-search-researcher to pull in external context when local documentation is insufficient.

What is the Comprehensive Research skill?

It is a specialized capability for Claude Code that automates deep technical research by coordinating multiple specialized agents across discovery, analysis, and synthesis phases.

What is the research confidence score?

After every investigation, the skill provides a rating from 0.0 to 1.0, indicating the reliability of the findings and highlighting any areas where evidence might be weak or missing.

How do I trigger this research workflow?

You can initiate the systematic investigation process by using the /ai-eng/research command within your Claude Code environment.

How does it handle large codebases?

It uses parallel discovery agents to map the codebase efficiently, followed by sequential analyzers that deep-dive into relevant files and patterns without losing context.

Comprehensive Research Orchestrator

Name: Comprehensive Research Orchestrator
Author: v1truv1us

byv1truv1us

•

开发者工具

Automates multi-phase codebase and documentation investigations to provide deep technical insights and evidence-based recommendations.

The Comprehensive Research skill coordinates specialized agents to perform deep-dive analysis of complex codebases, internal documentation, and external knowledge sources. By employing a systematic multi-phase workflow—Parallel Discovery, Sequential Analysis, and Synthesis—it ensures that developers make informed decisions based on a complete understanding of existing patterns, historical decisions, and architectural constraints. It features built-in confidence scoring and risk assessment to highlight potential gaps in knowledge, effectively preventing costly rework and architectural mistakes caused by incomplete discovery.

主要功能

01Evidence-based reporting with specific file and line-number references

02Automated identification of implementation patterns and historical constraints

032 GitHub stars

04Structured synthesis output with actionable immediate and long-term recommendations

05Multi-phase orchestration involving parallel discovery and sequential analysis

06Research Confidence Assessment scoring from 0.0 to 1.0

使用场景

01Synthesizing fragmented information from code, docs, and external sources into authoritative reports

02Investigating past design decisions and documentation before initiating major refactors or new features

03Exploring and documenting unfamiliar or legacy codebases to understand architectural dependencies

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add v1truv1us/ai-eng-system comprehensive-research

For use in Claude.ai and ChatGPT

Download Skill