deep-dive

Run a two-stage trace and interview workflow to produce a testable specification.

1|1|Updated Apr 13, 2026
One-click install
npx skills add https://github.com/Cheggin/request-for-startups --skill deep-dive-cheggin
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: deep-dive
Source: https://github.com/Cheggin/request-for-startups/tree/main/skills/deep-dive
Command: npx skills add https://github.com/Cheggin/request-for-startups --skill deep-dive-cheggin

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Deep Dive orchestrates a two-stage pipeline that first investigates why something happened (trace) and then precisely defines what to do about it (deep-interview). It uses three parallel trace lanes and a 3-point injection to seed the interview with trace-derived context, reducing redundant exploration and focusing questions on the most important unknowns.

Core Features & Use Cases

  • 2-stage trace + interview workflow with a structured injection mechanism that preserves context across phases
  • Parallel trace lanes that autonomously gather evidence for and against each hypothesis, then synthesize a ranked explanation
  • Output a standard deep-interview spec (.omc/specs/deep-dive-{slug}.md) ready for downstream planning and execution, including trace findings

Quick Start

Provide a problem statement and let the deep-dive orchestrator generate hypotheses, run the trace, inject results into the interview, and output a spec.

Frequently Asked Questions about deep-dive

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

FAQPage Schema
How do I translate ambiguous project requirements into a testable specification?

A deep-dive workflow crystallizes ambiguous requirements into a testable specification by first investigating root causes through parallel trace analysis, then conducting a focused interview to produce a formal spec for downstream planning.

What is the best way to uncover root causes from a vague problem statement?

Uncovering root causes from a vague problem statement requires parallel trace lanes that autonomously gather evidence for and against each hypothesis, synthesizing the findings into a ranked explanation before crystallizing requirements.

How does trace integration work when crystallizing requirements?

Trace integration works by using a 3-point injection mechanism to seed the interview phase with trace-derived context, preserving evidence across phases and focusing questions on the most important unknowns to produce a formal spec.

Can I use deep-interview specs for downstream planning and execution?

Yes, you can use deep-interview specs for downstream planning and execution because the workflow outputs a standard specification file at .omc/specs/deep-dive-{slug}.md that includes trace findings and is structured for direct consumption.

Do I need to manually generate hypotheses before starting a deep-dive trace?

No, you do not need to manually generate hypotheses before starting a deep-dive trace because the orchestrator automatically generates hypotheses, runs the trace, and injects results into the interview based on your initial problem statement.