engage-lexicon

Runs adversarial for/against/judge debate over contested legal, policy, and compliance claims under a declared jurisdiction.

Updated Mar 30, 2026
One-click install
npx skills add https://github.com/rubrical-works/idpf-praxis-skills --skill engage-lexicon-rubrical-works
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: engage-lexicon
Source: https://github.com/rubrical-works/idpf-praxis-skills/tree/main/Skills/engage-lexicon
Command: npx skills add https://github.com/rubrical-works/idpf-praxis-skills --skill engage-lexicon-rubrical-works

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires ajv, and includes references (resource) components.

What problem does it solve? Contested legal, policy, and compliance questions often get one-sided or jurisdiction-confused answers from a single-pass analysis. This Skill pressure-tests a directional legal claim by running a structured adversarial debate — a for-advocate, an against-advocate, and a judge — over live web research, with mandatory jurisdiction declaration and citation discipline distinguishing controlling from persuasive authority. ## Core Features & Use Cases - Adversarial debate workflow: Baseline pass, parallel for/against advocates, and a judge pass that endorses, rejects, or revises the claim with named weakening evidence and flip conditions. - Legal-domain enforcement: Mandatory jurisdiction gate, dual-axis citation diversity (zero URL overlap plus authority-hierarchy diversity), controlling-vs-persuasive annotation on every citation, and a statutory recency gate that flags superseded statutes and regulations. - Survey mode and safety contract: --mode survey refracts open-ended landscape questions across three analytical perspectives, and every output carries an educational-only disclaimer making clear it is not licensed legal advice. - Use Case: Ask whether California's AB-5 ABC test is preempted by FAAAA for motor carriers under US-CA-9th; the Skill debates both sides against Ninth Circuit authority, detects a stale AB-5 citation superseded by AB-2257, and returns a revised as-applied-preemption claim with medium confidence. ## Quick Start Invoke the skill with a directional legal claim and a declared jurisdiction, for example: run engage-lexicon with --jurisdiction US-DE-Chancery on the claim that MFW's dual-protection framework applies to this controller squeeze-out.

Frequently Asked Questions about engage-lexicon

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

FAQPage Schema
How do I analyze a contested legal claim with adversarial debate?▼

Invoke the skill with a directional claim and the --jurisdiction flag, for example --jurisdiction US-CA-9th. It runs a baseline pass, dispatches for and against advocates in parallel with live web research, then a judge pass that endorses, rejects, or revises the claim with cited controlling authority.

What is the difference between engage-lexicon, debate-prism, and engage-prism?▼

engage-lexicon handles directional legal, policy, and compliance claims under a declared jurisdiction. debate-prism covers business, market, and finance directional claims, while engage-prism handles open-ended exploration without a directional claim. engage-lexicon adds the jurisdiction gate, authority-hierarchy diversity, and statutory recency checks.

Does engage-lexicon work without Node.js installed?▼

Yes, it applies a no-runtime fallback pattern. Without Node.js 18+ and ajv, schema validation of advocate and judge JSON is skipped, but the full debate workflow, jurisdiction gate, citation diversity checks, and disclaimer stamping still run because they are enforced tool-side.

Can engage-lexicon replace a lawyer or provide legal advice?▼

No. Every output is stamped with an educational-only disclaimer stating it is not licensed legal advice and creates no attorney-client relationship. It produces a structured starting point for professional review and never drafts contracts, pleadings, or litigation strategy.

When should I use survey mode instead of debate mode?▼

Use --mode survey for open-ended landscape questions without a directional claim, such as how regulators are interpreting a new regulation. It dispatches three parallel perspective subagents instead of for/against advocates while keeping the jurisdiction gate, citation schema, and disclaimer.

Why does engage-lexicon halt when no jurisdiction is declared?▼

Legal claims are not jurisdiction-neutral, so the skill halts before claim extraction if --jurisdiction is missing. Without it, advocates could cite controlling authority from different forums and argue past each other, producing plausible-looking but meaningless analysis.