Our take
- Activity
- 28
- Community
- 44
- Issues
- 21
- Pull requests
- 43
+303 stars in the last 30 days
Project history
Daily GitHub snapshots since Oct 7, 2026. History covers stars and forks; health signals are not tracked historically.
Stars
10,025Forks
1,972Change since Nov 16, 2026: +9 stars·+1 forks
Overview
claude-for-legal is a collection of Claude plugins and managed-agent templates for recurring legal work. It targets in-house and firm practices, legal clinics, and law students, with workflows spanning contracts, corporate transactions, employment, privacy, product, regulation, AI governance, intellectual property, and litigation.
The project is most useful when a team wants configurable AI assistance grounded in its own playbooks, templates, and connectors rather than a general-purpose legal chatbot. It supports Claude Cowork and Claude Code plugins, as well as managed-agent deployment behind a workflow engine. Outputs are drafts and analysis aids, not legal advice or substitutes for attorney judgment.
Key Features
- Practice-area plugins with skills, slash commands, and practice profiles tailored through a cold-start interview.
- Contract, diligence, privacy, employment, regulatory, IP, litigation, clinic, and legal-learning workflows.
- Named agents for tasks such as agreement review, DSAR response, claim-chart building, and deadline monitoring.
- Managed-agent cookbooks and scripts for deploying scheduled workflows through the Claude Managed Agents API.
- MCP connectors for productivity tools, document and contract systems, e-discovery, dockets, and legal research platforms.
- Citation and review safeguards, including source attribution, surfaced jurisdiction assumptions, and gates before work is sent or filed.
- A community-skill hub with trust checks, license controls, updates, and skill quality review.
- Markdown and JSON configuration, with no build step described in the README.
Use Cases
- An in-house legal team can review vendor agreements against its playbook and route exceptions for approval.
- Privacy counsel can triage a new data use, prepare a PIA, or draft a DSAR response for attorney review.
- Corporate counsel can organize diligence findings, cited document reviews, and closing checklists during a transaction.
- Litigation teams can build chronologies, prepare deposition outlines, or monitor docket activity through configured workflows.
- A law clinic or student can use structured intake, memo scaffolds, case briefing, and study drills with appropriate supervision.
Getting Started
For Claude Code, add the repository as a marketplace and install a plugin:
/plugin marketplace add <path-to-this-repo-or-github-url>
/plugin install commercial-legal@claude-for-legal
Run that plugin's cold-start interview before using its other skills. See the README and QUICKSTART.md for Cowork setup, connector configuration, and managed-agent deployment.
Considerations
- The repository's outputs are drafts for attorney review. Users remain responsible for verifying sources, legal analysis, and final work product.
- Legal research connectors may require separate subscriptions, credentials, or API keys. Without a research tool, citations from model knowledge are flagged for verification.
- Skills are designed to work better after a cold-start interview and access to relevant playbooks or sample documents; without that setup, results may be generic.
- Managed-agent deployment requires an orchestrator and an Anthropic API key. The README identifies subagent delegation as a research-preview capability with a single delegation level.
- The repository is a broad set of reference templates, not a turnkey legal service. Teams should configure connectors, practice profiles, permissions, and review gates for their environment.
Found this useful?
Share it with someone who would like claude-for-legal.