Skip to content
← Tools / Skills
Skill · .md packageevolsb/10 trustVerified
Contract Review (evolsb) logo

Contract Review (evolsb)

AI contract review with CUAD risk detection and market benchmarks

Visit official site →

About

A contract-review skill grounded in the CUAD dataset's 41 legal risk categories from 510 real contracts, built after the author's own experience finding commercial legal AI overpriced and open-source alternatives incomplete. It gives position-aware review (flagging different risks depending on whether you are the buyer or seller), document-type checklists for NDAs, SaaS agreements, M&A and more, plus market-standard benchmarks and actual redline language. With 466 stars it is one of the most adopted skills in this batch — but the README is clear it is not legal advice and defaults to US law.

What it can do

  • Position-aware contract review
  • Red flags quick scan with severity ratings
  • Market standard benchmarks for key terms
  • Negotiability ratings
  • Jurisdiction awareness
  • M&A agreement support
  • Structured JSON redlines for deliverable generation

Setup

  1. Clone into your agent's skills directory:

git clone https://github.com/evolsb/claude-legal-skill ~/.claude/skills/contract-review

(Codex: ~/.codex/skills/contract-review)

  1. Ask the agent naturally, e.g.:

Review this NDA for red flags - I'm the receiving party

  1. Optional: pair with legal-redline-tools to generate tracked-changes Word docs and PDFs.

Prerequisites

Claude Code or another SKILL.md-compatible agent; contracts as files or pasted text.

Use cases

reviewing an NDA before signingbenchmarking an MSA against market standardspreparing redline language for negotiationscanning M&A agreements for rep survival and escrow terms

Works with

Claude CodeOpenAI CodexCursorGitHub CopilotGemini CLI

Trust score — 7/10

/10 trustVerified

Scored 7/10 (verified tier, metrics-v1). Evidence: 466 GitHub stars, 63 forks, 1 release at verification time (2026-10-13); highest-adoption legal skill in this batch.; Grounded in the CUAD dataset (41 categories, 510 contracts); sample outputs included.; Not legal advice; US-law default; companion legal-redline-tools repo for deliverables.; MIT license confirmed on GitHub license field.. License: MIT License, verified via GitHub license field..

★ 466 GitHub stars

How the trust score is built

Full methodology →
Adoption · stars/downloads, log-scaled vs peers
7.0/10
Maintenance · recency, releases, issue load
6.0/10
Openness · license + docs
8.0/10
Manual review · sanity read + red-flag check
7.0/10
  • · 466 GitHub stars, 63 forks, 1 release at verification time (2026-10-13); highest-adoption legal skill in this batch.
  • · Grounded in the CUAD dataset (41 categories, 510 contracts); sample outputs included.
  • · Not legal advice; US-law default; companion legal-redline-tools repo for deliverables.
  • · MIT license confirmed on GitHub license field.

Sources

Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.