rule-arc

Convert craft observations into structured rule-arcs with provenance and tier labels.

1|Updated Feb 19, 2026

Search Tags:#registry #prompt-engineering #llm-testing #doctrine #rule-arc #craft-rule #bite-test

Ready to run

npx skills add https://github.com/mrrts/WorldThreads --skill rule-arc

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: rule-arc
Source: https://github.com/mrrts/WorldThreads/tree/main/.claude/skills/rule-arc
Command: npx skills add https://github.com/mrrts/WorldThreads --skill rule-arc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Converts lived craft-shape observations into a structured, auditable rule-arc that carries provenance, bite-tests, and tier labels.

Core Features & Use Cases

Discipline-driven progress: turns observations into a documented arc with provenance and test results.
Tiered evaluation workflow: supports Unverified through EnsembleVacuous with cross-character context.
Documentation triangle integration: updates CLAUDE.md and OBSERVATIONS.md as the arc progresses.

Quick Start

Supply a lived craft observation and let this skill guide you through the bite-test arc, tier assignment, and accompanying documentation.

Frequently Asked Questions about rule-arc

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

FAQPage Schema

How do I turn prompt-engineering observations into verified rules with provenance?▼

A rule-arc is a structured, auditable progression that converts lived craft-shape observations into test-backed rules with provenance and tier labels. It enforces disciplined progress by requiring frontmatter and documentation updates as the arc completes.

How do I apply bite-tests to evaluate LLM-testing observations?▼

You apply bite-tests by supplying a lived craft observation, which the skill uses to guide you through the bite-test arc, tier assignment from Unverified to EnsembleVacuous, and cross-character context evaluation while updating CLAUDE.md and OBSERVATIONS.md.

Can I use cross-character probes to document registry entries into a rule-arc?▼

Yes, you can use cross-character probes to document registry entries. The skill applies to lived-play observations, registry entries, and cross-character probes to produce documented, test-backed rule progress with provenance and tier labels.

What is a rule-arc and how does it structure craft observations?▼

A rule-arc structures craft observations by converting them into an auditable progression with provenance, bite-tests, and tier labels. It enforces a disciplined workflow requiring a frontmatter name and description in SKILL.md and updates to CLAUDE.md and OBSERVATIONS.

Does the rule-arc workflow require specific frontmatter in SKILL.md?▼

Yes, the rule-arc workflow enforces discipline by requiring a frontmatter name and description in SKILL.md. It also updates CLAUDE.md and OBSERVATIONS as the arc completes to maintain documentation triangle integration.

What is the best way to document tiered evaluation progress for LLM-testing rules?▼

The best way to document tiered evaluation progress is to use a structured rule-arc that supports a tiered evaluation workflow from Unverified through EnsembleVacuous with cross-character context, automatically updating CLAUDE.md and OBSERVATIONS as the arc progresses.