Anthropic’s Claude For Legal platform now ships over 90 named AI agents on GitHub, each designed to run a specific legal workflow with a single command. The agents cover contract review, privacy compliance, litigation support, corporate due diligence, and law school clinic work. Separately, AmLaw 200 firm Hanson Bridgett announced firm-wide adoption of Claude, becoming the second Big Law firm after Freshfields to publicly commit to the platform.
What the Agents Actually Do
The 90+ agents go well beyond the 12 core plugins and MCP connectors that dominated coverage when Claude For Legal formally launched. Each agent has a job-style name tied to its workflow: Vendor Agreement Reviewer, NDA Triager, DSAR Responder, Amendment Tracer, Termination Reviewer, Claim Chart Builder, according to the GitHub repository.
Some agents are interactive, running on demand when a lawyer invokes them. Others are scheduled: the Renewal Watcher scans contract registers for cancel-by deadlines, the Deal Debrief runs weekly sweeps of signed agreements flagging playbook deviations, and the Playbook Monitor watches deviation logs and proposes updates when clause language has drifted, per Artificial Lawyer.
Every agent is customizable in natural language. Lawyers can modify agent behavior, practice profiles, and connector configurations without writing code. The agents run on Claude Cowork, Claude Code, or through the Claude Managed Agents API for teams that want to integrate them into existing workflow engines.
Mark Pike, associate general counsel at Anthropic and the lead on the Claude For Legal rollout, told Artificial Lawyer that the platform is designed around verification, not replacement: “Source attribution on citations, jurisdiction established during an onboarding interview, and explicit gates before anything is filed, sent, or relied on. The lawyer reviews and verifies; the tooling is designed to make that review easier, never to skip it.”
Hanson Bridgett Goes Firm-Wide
The same day, San Francisco-based Hanson Bridgett (approximately 200 lawyers) announced it had adopted Claude firm-wide for attorneys and professional staff, including the legal add-ons. The firm is using the platform for document review, drafting, deposition summary, version comparison, and due diligence in corporate transactions, according to Artificial Lawyer.
Laura Long, COO and CFO at Hanson Bridgett, said the firm sees AI “as an opportunity to lead responsibly, support our attorneys and professional staff, and continue building a more thoughtful and forward-looking firm.” The firm has a written AI use policy with restrictions on what information enters AI systems, plus enterprise-grade data protections.
The Vertical Agent Distribution Model
The 90+ agent count matters less than what it represents: a model vendor shipping pre-built, customizable agents for a specific profession as a distribution strategy. Anthropic is not selling a general-purpose API and hoping lawyers figure it out. It is packaging Claude into named workflows that match how legal teams already describe their work.
The competitive landscape is not empty. Harvey, Legora, Lexis, and Thomson Reuters all offer agent and workflow customization, and they have the advantage of model flexibility. Claude For Legal locks users into Anthropic’s models. But the Apache 2.0 open-source license and GitHub-first distribution lower the barrier for firms that want to experiment before committing to a vendor.
Two Big Law firms publicly declaring for Claude in the span of weeks, Freshfields and now Hanson Bridgett, suggests the vertical agent distribution model is gaining traction with exactly the kind of risk-averse, compliance-heavy buyers that enterprise AI vendors have struggled to reach.