deep-analysis

Analyze a prepared task across product, system, and constraints/risk streams with critic validation.

Updated Mar 16, 2026
One-click install
npx skills add https://github.com/nolood/planpipe --skill deep-analysis-nolood
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: deep-analysis
Source: https://github.com/nolood/planpipe/tree/main/skills/deep-analysis
Command: npx skills add https://github.com/nolood/planpipe --skill deep-analysis-nolood

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill provides a structured, multi-stream investigation of a task that has already been prepared or scoped, done BEFORE plan creation. It ensures thorough understanding and knowledge base building for the next stage (plan synthesis).

Core Features & Use Cases

  • Parallel Analysis Streams: Concurrently analyzes the task from product, system, and constraints/risk perspectives.
  • Analysis Critique: Reviews and validates the analyses produced by subagents.
  • Artifact Assembly: Assembles detailed analysis artifacts for the next stage.
  • User Reporting: Provides a summary and options for proceeding to Stage 3.

Quick Start

Invoke the deep-analysis skill with the appropriate command, such as: deep-analysis analyze task123.

Frequently Asked Questions about deep-analysis

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

FAQPage Schema
How do I perform a deep task analysis before plan creation?▼

Deep task analysis uses three concurrent streams to investigate product, system, and constraint perspectives. This multi-stream process builds a thorough knowledge base and assembles analysis artifacts before you synthesize a plan.

What is parallel analysis for software engineering tasks?▼

Parallel analysis concurrently evaluates a prepared task from product, system, and risk perspectives. An independent critic validates these subagent analyses to ensure thorough understanding before artifact assembly and downstream reporting.

Do I need a prepared task input to run deep analysis?▼

Yes, deep analysis requires input from Stage 1 with a READY_FOR_DEEP_ANALYSIS verdict. You must also provide access to the task's relevant codebase and documentation to conduct the investigation.

How does analysis validation work with concurrent streams?▼

Analysis validation uses an independent critic to review and validate the analyses produced by subagents. This ensures the product, system, and constraints/risk perspectives are accurate before assembling the final artifacts.

When do I need multi-stream investigation for task analysis?▼

You need multi-stream investigation when a task is scoped but lacks a comprehensive knowledge base. It provides structured analysis of product, system, and risk factors required before moving to plan synthesis.

Are there limitations to using concurrent analysis streams for task preparation?▼

The process is limited to tasks that have already been prepared and scoped. It cannot run without a Stage 1 READY_FOR_DEEP_ANALYSIS verdict and access to the relevant codebase and documentation.