Initialize Ralph Research Project

Create a new Ralph loop project directory with customized spec and progress files.

Usage

• •/init-ralph-research <topic> - Create project for the given topic
• •/init-ralph-research - Will prompt for topic

Reference Files

• •@spec-template.md - Template for spec.md with v2 features (dynamic tasks, deep research workflow)
• •@progress-template.md - Template for progress.md with questions section
• •@assumptions-template.md - Template and guidance for assumptions-and-questions.md

Instructions

1. •

   Get project topic:

   • •If $ARGUMENTS provided, use that as the topic
   • •Otherwise, ask: "What topic should this research project cover?"
2. •

   Derive project name:

   • •Convert topic to kebab-case for directory name (e.g., "Monte Carlo Simulation" → "monte-carlo-simulation")
   • •Keep it concise (3-4 words max)
3. •

   Create project directory:

   bash

   mkdir -p docs/claude-learnings/<project-name>
   

4. •

   Create spec.md using @spec-template.md:

   • •Replace <PROJECT_NAME> with the topic (title case)
   • •Replace <TOPIC> with the topic
   • •Replace <output_file> with info.md
   • •Adjust References section to use correct relative path to repository root
5. •

   Create progress.md using @progress-template.md:

   • •Replace <TOPIC> with the topic
6. •

   Confirm to user:

   code

   Created Ralph research project: docs/claude-learnings/<project-name>/
   
   Files created:
   - spec.md
   - progress.md
   
   To run the loop:
   bash .claude/lab/ralph/wiggum.sh docs/claude-learnings/<project-name> 10
   

Example

/init-ralph-research options pricing models

Created Ralph research project: docs/claude-learnings/options-pricing-models/

Files created:
- spec.md (configured for "Options Pricing Models" research)
- progress.md (ready for first iteration)

To run the loop:
bash .claude/lab/ralph/wiggum.sh docs/claude-learnings/options-pricing-models 10

Research Conventions

info.md as Investigation Tracker

The "Areas for Deeper Investigation" section in info.md serves as the tracker for research progress. Use strikethrough + links to show completed investigations:

## Areas for Deeper Investigation

1. ~~Token optimization strategies~~ → See [token-optimization.md](./token-optimization.md)
2. ~~CI/CD integration~~ → See [ci-cd-integration.md](./ci-cd-integration.md)
3. Quality metrics - measuring output quality across iterations
4. Context handoff patterns - efficient state transfer between iterations

Why this format:

• •Visual progress: Strikethrough shows what's done at a glance
• •Navigation: Links let readers jump to detailed research
• •Task generation: Remaining non-struck items drive new Deep Research tasks

Alternative for lighter investigations (no separate file produced):

- [x] Token optimization strategies
- [ ] Quality metrics

Deep Research: Separate Files vs. Consolidation

Create separate <topic>.md files when:

• •Research exceeds ~200 lines
• •Topic is self-contained with its own sources
• •Content includes code samples, tables, or detailed specifications
• •Topic may be referenced independently

Append to info.md when:

• •Research is brief (<100 lines)
• •Findings are closely tied to core research
• •Topic doesn't warrant standalone navigation

The hybrid approach keeps info.md lean as a navigation hub while substantial research lives in dedicated files.