Claude Code is a command-line AI assistant that transforms developer workflows by integrating with your existing tools. Using Anthropic’s state-of-the-art Claude 4 models through AWS Bedrock, it creates a unified, context-aware workspace for coding, documentation, and project management.

Setup: Claude Code with AWS Bedrock

Prerequisites

  1. Install Claude CLI: Follow the official quickstart guide
  2. AWS Authentication: Access to cambri-datascience-dev account with proper credentials

Configuration

Set these environment variables in your .bashrc or .zshrc:

# Enable Bedrock integration
export CLAUDE_CODE_USE_BEDROCK=1
export ANTHROPIC_MODEL="eu.anthropic.claude-sonnet-4-20250514-v1:0"
export AWS_DEFAULT_REGION="eu-west-1"

# AWS Profile setup
export AWS_PROFILE="DSDEV"

Authentication

# Login to AWS SSO
aws sso login

# Launch Claude Code
claude

Test: Say “hi” to verify the connection (avoid “test” as it may trigger unit test generation).

MCP Server Integration

Extend Claude’s capabilities by connecting to development tools:

ServiceMCP Server Package
Jira@aashari/mcp-server-atlassian-jira
Confluence@aashari/mcp-server-atlassian-confluence
GitHub@missionsquad/mcp-github
Slack@modelcontextprotocol/server-slack
AWS Docsawslabs.aws-documentation-mcp-server@latest

Key Use Cases

1. Code Development

  • Feature Implementation: “Implement the API described in Jira ticket ENG-123 and create a PR”
  • Code Reviews: Automated security scanning and vulnerability detection on pull requests
  • Documentation: Auto-sync code changes with README and Confluence pages

2. Project Management

  • Ticket Queries: “What’s the status of our current sprint?”
  • Bulk Operations: Create multiple sub-tasks from requirements automatically
  • Sprint Reports: Generate and distribute sprint summaries via Slack

3. Communication & Knowledge

  • Slack Integration: Query Jira/Confluence directly from team chat
  • Document Search: “What are our password complexity requirements?”
  • Debugging: Cross-reference error logs, code, and documentation for root cause analysis

IDE Integration

Claude Code integrates with popular development environments:

  • JetBrains IDEs: PyCharm, WebStorm, IntelliJ IDEA via official plugin
  • VS Code: Auto-installs when running claude from the terminal
  • Terminal: Full CLI interface for any development setup

Security & Data Handling

  • Local Processing: Runs entirely in your terminal, no remote code indexing
  • AWS Bedrock: Models invoked through existing AWS data agreements
  • Direct API: No third-party backend servers required

Example Workflow

  1. Requirements: Jira story describes new feature
  2. Implementation: “Claude, implement this feature”
  3. Testing: Auto-generated unit tests
  4. Review: AI security scan on PR
  5. Documentation: Update Confluence with feature details
  6. Communication: Notify team in Slack

All through natural language, eliminating context switching between tools.

Benefits

  • Reduced Context Switching: Access all tools through one interface
  • Automated Security: AI catches vulnerabilities before merge
  • Faster Debugging: Multi-source error analysis
  • Living Documentation: Code and docs stay in sync
  • Team Communication: Project data accessible via Slack

Getting Started

Claude Code + MCP transforms development into a conversational workflow. Instead of juggling multiple tools, developers interact with their entire tech stack through natural language, creating a truly context-aware development environment.

Experience from implementing AI-powered development workflows with Claude Code and AWS Bedrock.