Anthropic released 10 pre-built agent templates targeting the most labor-intensive workflows in financial services: pitchbook creation, KYC file screening, general ledger reconciliation, and month-end close. Each template ships as a plugin for Claude Cowork and Claude Code, and as a cookbook for Claude Managed Agents, with the full set available on GitHub. Bloomberg first reported the release on May 5.
What the Templates Cover
The 10 agents split across two categories, according to Anthropic’s announcement.
Research and client coverage includes five agents: a pitch builder that creates target lists, runs comparables, and drafts pitchbooks; a meeting preparer that assembles client and counterparty briefs; an earnings reviewer that reads transcripts, updates models, and flags thesis-relevant changes; a model builder for maintaining financial models from filings and data feeds; and a market researcher that tracks sector developments across news, filings, and broker research.
Finance and operations includes five more: a valuation reviewer that checks against comparables and firm review standards, a GL reconciler for general ledger accounts and NAV calculations, a month-end closer that runs the close checklist and produces journal entries, a statement auditor for consistency and audit-readiness checks, and a KYC screener that assembles entity files and packages escalations for compliance review.
Two Deployment Models
The templates run in two modes. As a Claude Cowork or Claude Code plugin, the agent works alongside an analyst on their desktop. Anthropic’s example: hand the pitch agent a target list, and it returns a comps model in Excel, a pitchbook drafted in PowerPoint, and a cover note ready in Outlook.
As a Claude Managed Agent, the same template runs autonomously on Claude’s platform for batch workloads. The cookbooks include long-running sessions for multi-hour deal closes, per-tool permissions, managed credential vaults, and a full audit log in Claude Console where compliance teams can inspect every tool call and decision. In both cases, humans review and approve output before it reaches clients or gets filed.
Data Connector Expansion
Alongside the templates, Anthropic announced eight new data connectors and one MCP app. New connectors include Dun & Bradstreet for verified business identity, Fiscal AI for real-time equity fundamentals, Guidepoint for compliance-reviewed expert interview transcripts, SS&C Intralinks for DealCenter AI data rooms, Third Bridge for primary-source expert interviews, and Verisk for insurance underwriting data. Moody’s launched an MCP app providing proprietary credit ratings and data on more than 600 million public and private companies.
These join existing connectors for FactSet, S&P Capital IQ, MSCI, PitchBook, Morningstar, Chronograph, LSEG, and Daloopa.
The Template Distribution Model
The release reflects a broader shift in how AI labs are selling agent capabilities to regulated industries. Rather than offering raw API access and expecting financial institutions to build their own agents from scratch, Anthropic is packaging domain-specific workflows with built-in governance controls. ABA Banking Journal noted that the competitive prize in agent adoption “may go to the AI app or suite of tools that integrates best into the established technology stack relied upon by financial services, advisory, and wealth management teams.”
Anthropic says the templates pair best with Claude Opus 4.7, which leads Vals AI’s Finance Agent benchmark at 64.37%. The templates also arrive weeks after Anthropic and FIS launched a separate Financial Crimes AI Agent for AML investigations, and after the company announced a $1.5 billion enterprise AI services joint venture with Blackstone, Goldman Sachs, and other firms.
For wealth management teams that have been experimenting with meeting prep chatbots and document summarizers, the templates represent a step change in scope: agents that can run an entire month-end close or build a pitchbook from a target list, with audit trails attached. Whether financial institutions trust them to do so at scale is the open question.