ralph
Specification-first AI development powered by Ouroboros. Socratic questioning exposes hidden assumptions before writing code. Evolutionary loop (Interview → Seed → Execute → Evaluate → Evolve) runs until ontology converges. Ralph mode persists until verification passes — the boulder never stops. Use when user says "ralph", "ooo", "don't stop", "must complete", "until it works", "keep going", "interview me", or "stop prompting".
83
83%
Does it follow best practices?
Impact
71%
2.02xAverage score across 3 eval scenarios
Advisory
Suggest reviewing before use
Evaluation results
40%
33%Fix All Failing Tests in a Legacy Python Module
Ralph state file and loop reporting
State file location
0%
0%
State file mode field
0%
0%
State file session_id
0%
0%
State file required fields
0%
0%
Default max iterations
0%
0%
Iteration progress header
0%
75%
Verification result in report
0%
25%
Score in report
0%
100%
Boulder never stops phrase
0%
100%
Verification history populated
0%
0%
Checkpoint saves
0%
0%
Ralph COMPLETE header
0%
71%
Completion report fields
100%
100%
Verification history in completion
0%
100%
100%
57%Design a Personal Budget Tracker CLI
Specification interview and seed generation
Interview questions produced
100%
100%
Ambiguity score computed
40%
100%
Ambiguity formula applied
0%
100%
Greenfield vs brownfield weights
0%
100%
Threshold check
30%
100%
YAML seed format
100%
100%
Seed goal field
28%
100%
Seed constraints field
28%
100%
Seed acceptance_criteria field
0%
100%
Seed ontology_schema field
10%
100%
No code before spec
100%
100%
Spec saved to file
100%
100%
74%
18%Evolve an Ontology for a Logistics Event System
Evolutionary loop convergence tracking
Multiple generations produced
100%
100%
Ontology schema per generation
100%
100%
Similarity formula applied
0%
100%
Convergence threshold
80%
100%
Stagnation detection: consecutive
0%
0%
Stagnation detection: oscillation
0%
0%
Hard cap 30 generations
0%
0%
Per-generation similarity score
100%
100%
Wonder-Reflect-Seed pattern
12%
62%
Lineage tracking
100%
100%
Convergence action reported
100%
100%
No premature convergence
100%
100%
- Repository
- supercent-io/skills-template
- Commit
c033769
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.