Skip to content
HomeNewsDigestsConceptsGuidesToolbox
AboutSubscribeUA
Subscribe

AI Today Brief

The daily AI-engineering brief. Built in public. EN · UA.

XTelegramLinkedInYouTubeRSS

Follow AI Today Brief on LinkedIn for daily AI-engineering updates and the weekly “5 shifts that changed how developers work” PDF.

Explore

NewsDigestsConceptsGuides

Company

SubscribeAdvertiseAbout

Legal

Editorial policyAI disclosurePrivacyTerms

© 2026 AI Today Brief. All rights reserved.

  1. Home/
  2. News/
  3. Agents & MCP/
  4. Anthropic Open-Sources Financial Reference Agents and Model Context Protocol Connectors
Agents & MCP

Anthropic Open-Sources Financial Reference Agents and Model Context Protocol Connectors

Anthropic published an open-source suite of reference workflow agents and Model Context Protocol data connectors. The repository provides ready-to-run slash commands and dual deployment for Claude Cowork and headless APIs.

September 25, 2026· 5 min read
OKCurated by Oleksandr Kuzmenko, AI Product Engineer·Updated September 25, 2026·Sources cited on every story
AI-assisted · editor-reviewed·How we use AI
Anthropic Open-Sources Financial Reference Agents and Model Context Protocol Connectors

Impact: Medium

Why it matters

You can inspect and deploy Anthropic's reference agent architecture, slash command definitions, and data connectors directly into Claude Code or your custom backend.

TL;DR

  • 01Anthropic open-sourced modular agent patterns runnable via Cowork plugins or the /v1/agents API.
  • 02Centralizes MCP configurations for commercial financial data feeds inside standard JSON schemas.
  • 03Provides an administrative Claude Code plugin to provision Microsoft 365 add-ins against private cloud gateways.

Key facts

Format
File-based markdown and JSON, zero build steps
API Endpoint
Claude Managed Agents (/v1/agents)
Cloud Routing Support
Direct API, AWS Bedrock, Google Vertex AI

Dual Deployment Architecture

Anthropic's anthropics/financial-services repository provides production-grade templates for specialized agent workflows. Every workflow ships from a single file-based layout that requires no build step, deploying identically via two distribution targets:

1. As a desktop plugin installed inside Claude Cowork. 2. As an autonomous backend workflow provisioned via the Claude Managed Agents API (/v1/agents).

Slash Commands and Model Context Protocol Integrations

Workflows incorporate dedicated slash commands (/comps, /dcf, /earnings, /ic-memo) configured alongside Model Context Protocol connectors. The centralized financial-analysis plugin bundles shared financial modeling skills with MCP configurations for data providers like S&P Capital IQ, Moody's, and MT Newswires.

Subagent orchestration is implemented through a reference event loop in scripts/orchestrate.py, which manages task delegation to depth-1 leaf workers via callable_agents and routes handoff_request events between specialized runtimes.

Enterprise Microsoft 365 Provisioning

The repository also includes administrative automation under claude-for-msft-365-install/. This Claude Code plugin assists IT operators in generating customized manifests to route Office add-in requests through AWS Bedrock, Google Vertex AI, or internal private enterprise gateways instead of the default Anthropic API endpoints.

Try it in 2 minutes

claude plugin marketplace add anthropics/financial-services
claude plugin install financial-analysis@claude-for-financial-services

bash

✓ When to use

  • Architecting complex multi-agent systems that need clear subagent delegation patterns.
  • Connecting Claude or Claude Code directly to institutional financial data providers via Model Context Protocol.

✕ When NOT to use

  • Automated financial operations requiring autonomous execution without human verification or sign-off.
  • Simple coding assistant tasks where generalist Claude Code out-of-the-box suffices.

What to do today

  • →Add the repository to your Claude plugin marketplace using claude plugin marketplace add anthropics/financial-services.
  • →Review scripts/orchestrate.py for Anthropic's reference implementation of subagent delegation and handoffs.
  • →Inspect plugins/vertical-plugins/financial-analysis/.mcp.json to model your own external MCP connections.
#Claude#Claude Code#Model Context Protocol#Microsoft 365

Sources

  • Claude for Financial Services GitHub Repository
ShareShare on XShare on LinkedIn
← Previous storyWhiteboard Open-Sources Visual Design Integrated Development Environment for Coding AgentsNext story →Benchmarking Claude Opus 5.5: Why Agent Skills and CLI Updates Still Matter

Related stories

  • Agents & MCPCursor Shares Harness Tuning Prompt to Cut Agent Token Overhead
  • Agents & MCPBenchmarking Claude Opus 5.5: Why Agent Skills and CLI Updates Still Matter
  • Agents & MCPProduction Readiness Checklist for Hardening Model Context Protocol Servers
  • Agents & MCPAnthropic and WHO Automate Complex Outbreak Reporting via Claude Skills

Email digest

Get the morning AI brief

One email a day — the stories that matter for engineers, founders and tech leads. Human-edited, with links to primary sources.

  • ✓120+ sources scanned daily
  • ✓Edited by a human
  • ✓1 email per day
  • ✓EN + UA

By subscribing you agree to the privacy policy.