Review contracts like NDAs and SaaS agreements in Claude Code
The author built an open-source contract-review skill for Claude Code and other AI coding tools to review NDAs, SaaS agreements, vendor terms and merchant agreements. Task: Suggested implementation: Package a contract-review skill as an open-source Claude Code add-on, installed via GitHub into the local skills folder. Use it to review NDAs, SaaS agreements, vendor terms, and merchant agreements. Human role: Verify legal accuracy and risk flags against standard templates.; tools: Claude Code. Human role: The original post does not say how a person stayed involved.; limitations: Prerequisites include local git access and Claude Code installation. Integration uncertainty exists regarding compatibility with other AI coding tools. Source reliability is low due to HTTP 403 error and lack of detailed execution logs. Human approval is mandatory before relying on outputs for legal decisions..
Tools
Published on Agent Work Atlas:
Overview
The author built an open-source contract-review skill for Claude Code and other AI coding tools to review NDAs, SaaS agreements, vendor terms and merchant agreements.
Workflow
Suggested implementation: Package a contract-review skill as an open-source Claude Code add-on, installed via GitHub into the local skills folder. Use it to review NDAs, SaaS agreements, vendor terms, and merchant agreements. Human role: Verify legal accuracy and risk flags against standard templates.
Reported result
Actual outcomes are unknown as the source excerpt provides no performance metrics or success data. Proposed acceptance check: Measure time saved per contract review compared to manual process and verify zero critical legal errors in a sample set.
Human role
The original post does not say how a person stayed involved.
Limitations
Prerequisites include local git access and Claude Code installation. Integration uncertainty exists regarding compatibility with other AI coding tools. Source reliability is low due to HTTP 403 error and lack of detailed execution logs. Human approval is mandatory before relying on outputs for legal decisions.
Try it
I need to pilot a contract review workflow using Claude Code. First, clone the open-source contract-review skill from [replaceable GitHub URL] into ~/.claude/skills/contract-review. Next, select a sample NDA or SaaS agreement from [replaceable file path]. Run the skill to generate a review report. Then, manually compare the AI's risk flags against our standard legal templates to verify accuracy. Do not publish or act on any findings without my explicit confirmation. If the skill fails to install or produces no output, stop and report the error. Measure the time taken for this review versus our manual baseline. Ask me for clarification if the sample contract type is ambiguous.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.