Config Authentication

Step 1: Clone the .env.template and rename to .env Step 2: Run command

chainlit create-secret

Step 3: Copy the chainlit secret into .env and update other environment variables

Config Data Persistent

Step 1: Update POSTGRES_xxx variables in .env file correctly Step 2: Start postgres container by using command

docker-compose up -d

Step 3: Restart server to apply tables to database

Real-Time Log Streaming (NEW ⚡)

The system now supports real-time log streaming from workers to the hub via Redis Pub/Sub and PostgreSQL persistence:

• ✅ Push-based delivery - No polling overhead
• ✅ Persistent storage - Jobs survive restarts
• ✅ Sub-100ms latency - Real-time UI updates
• ✅ Automatic fallback - Works with or without Redis

Quick Setup

# 1. Start Redis
docker-compose up -d redis

# 2. Install dependencies
make setup

# 3. Initialize database tables
make init-streaming

# 4. Configure environment (.env)
REDIS_URL=redis://localhost:6379
ENABLE_LOG_STREAMING=true

📖 See STREAMING_MIGRATION_GUIDE.md for detailed setup instructions.

MCP Server Integration

The Chainlit agent expects an MCP server that exposes its tools over HTTP:

• Default endpoint: http://localhost:1337/mcp
• Configure with the MCP_SERVER_URL environment variable
• Optional timeout override: MCP_HTTP_TIMEOUT (seconds)

MCP servers are now hosted in the separate sample-mcp-servers repository:

Available MCP Servers

1. Code Scout MCP Server

Fast synchronous symbol scanner for Python codebases.

• Repository: sample-mcp-servers/mcp_servers/code-scout/
• Tools: scan_directory, find_symbol, analyze_impact, grep_search, git_blame, build_dependency_graph
• Supports local paths and GitHub URLs
• Set GITHUB_TOKEN for private repositories

2. Refactoring Agent MCP Server

AI-powered code refactoring using CrewAI and AWS Bedrock with real-time log streaming.

• Repository: sample-mcp-servers/mcp_servers/refactoring-agent/
• Tools: start_refactoring_job, get_job_status
• Requires AWS credentials with Bedrock access
• Feature: Real-time progress via Redis Pub/Sub + PostgreSQL persistence

See the sample-mcp-servers README for detailed setup and usage instructions.