Laboratory Patterns Skill

Scientific experimentation patterns for exploring implementation approaches.

Overview

This skill provides patterns and templates for the /serena-refactor:laboratory command, which applies scientific methodology to code implementation:

• Problem Decomposition: Breaking complex problems into testable units
• Approach Collection: Gathering solutions from multiple sources
• Creative Generation: Novel approaches beyond documentation
• Hypothesis Formation: Structuring testable experiments
• Parallel Experimentation: Running multiple PoCs concurrently
• Rigorous Verification: Testing with edge cases and reproducibility
• Learning from Failure: Recording what doesn't work and why

Core Philosophy

"Apply scientific methodology to code implementation"

1. Decompose - Break problems into smaller, testable units
2. Hypothesize - Form clear, testable hypotheses
3. Experiment - Run controlled experiments (PoCs)
4. Verify - Test rigorously with edge cases
5. Learn - Record successes AND failures

References

7-Stage Workflow

1. Problem Decomposition  [User Approval]
      ↓
2. Existing Approaches Collection
      ↓
3. Creative Approaches Generation
      ↓
4. Hypothesis Formation  [User Approval]
      ↓
5. PoC Experiments (Parallel, max 5)  [User Approval]
      ↓
6. Verification
      ↓
7. Result Report

Agent Roles

Information Sources (Priority)

1. GitHub Issues/Discussions - Real developer solutions
2. Stack Overflow - Community-vetted approaches
3. Official Docs (Context7) - Standard methods

Verification Criteria

State Files

Output Locations

• Reports: docs/lab-reports/YYYY-MM-DD-{topic}.md
• Memory: .serena/memories/lab-{topic}.md
• Experiments: experiments/hypothesis-{id}/