Skip to content

Legal Blueprint

Review Legal Contracts With AI-Powered Analysis

evolsbMay 2026

A single skill that analyzes contracts using CUAD risk categories, market benchmarks, and position-aware recommendations.

Contract Review

Install Command

Run this command to deploy the blueprint to your environment.

What problem does this solve?

Analyzes legal contracts and surfaces unfavorable terms, missing provisions, and jurisdiction risks in a single structured review. This removes the need to manually cross-reference clauses against market standards or guess which terms are actually negotiable.

How does it work?

  1. Upload or paste a contract and specify which party you represent (customer, vendor, buyer, seller, etc.). 2. The skill identifies document type, runs red flag detection, and analyzes each risk category from CUAD. 3. You receive a structured markdown report with severity ratings, market benchmarks, negotiability scores, and specific redline language. Output: a lawyer-ready review with ranked negotiation priorities.

Included Skill Groups

What's the biggest win?

Material risks surface with market context and actual redline language, so you enter negotiations informed instead of guessing what to ask for.

What should I know technically?

Built on the CUAD dataset (41 legal risk categories from 510 real contracts), LegalBench, and ContractEval benchmarks. Requires Claude or compatible AI assistant (Claude Code, OpenAI Codex, Cursor, GitHub Copilot, Gemini CLI, or 26+ Agent Skills-compatible tools). No external API keys or databases required beyond your AI assistant.

What should I watch out for?

Not a substitute for attorney review on material deals. US law focused (analysis varies by jurisdiction). Very long contracts may require section-by-section review due to context window limits. F1 score ~0.62 on clause extraction per ContractEval benchmarks, so best used for first-pass review and issue flagging.

Key Features

Position-Aware Review

Analysis adjusts severity based on which party you represent (customer vs. vendor, buyer vs. seller, etc.), so you see what is actually risky for YOUR position

Market Standard Benchmarks

Compares contract terms against industry norms with clear thresholds (e.g. liability cap 12 months = standard, 6 months = yellow flag, under 6 = critical)

Red Flags Quick Scan

Instant detection of 20+ danger signs (uncapped indemnification, unilateral amendment rights, perpetual obligations, offshore jurisdiction) before deep analysis

Document-Type Checklists

Specialized checklists for NDAs, SaaS/MSAs, M&A agreements, payment/merchant agreements, and finder/broker agreements with type-specific risks

Specific Redline Language

Outputs actual replacement text, not just flagged issues. Includes fallback positions and negotiability ratings so you know what is realistic to ask for

Tools in this Blueprint

Claude logo
4.7(315 reviews)
OpenAI Codex
Cursor
GitHub

About This Blueprint

License
MIT
Industry
Technology
Skills
0 workflows, 0 sub-skills, 1 standalone