FlameFlare Worker Examples

This directory contains real-world Cloudflare Worker examples that demonstrate various patterns and features supported by FlameFlare.

Available Examples

JavaScript/TypeScript Examples

Example	Description	Features
hello-world/	Simplest possible worker	Basic Response, zero config
json-api/	REST API with routing	JSON responses, URL routing, HTTP methods
environment-variables/	Using env vars and secrets	Environment bindings, secret management
router-worker/	URL routing with CORS	Advanced routing, CORS, HTML/JSON responses
scheduled-worker/	Cron trigger patterns	Scheduled events, dual handlers
multi-module/	Multi-file worker with imports	ES module imports, JSON modules, code splitting
wasm-module/	WASM module import	WebAssembly import, binary modules, WebAssembly.instantiate()
queue-producer-consumer/	Message queue patterns	Queues, producers, consumers, batch processing

Workflow Examples

Example	Description	Features
workflow-basic/	Multi-step document signing	Durable steps, sleep, retry config, status tracking
workflow-events/	Event-driven payment verification	waitForEvent, sendEvent, webhook patterns
workflow-comprehensive/	User onboarding pipeline (all features)	Every Workflow API: parallel steps, sleepUntil, conditional branching, dynamic loops, try/catch, NonRetryableError, all instance management

Integration Examples

Example	Description	Features
order-processing-pipeline/	Full e-commerce pipeline (Workers + Queues + Workflows)	All binding types, queue producer/consumer chains, workflow with event waiting, cron triggers, service bindings, multi-worker coordination

AI Multi-Agent Examples

Example	Description	Features
ai-code-review-council/	Multi-agent AI code review with BAML	3 specialist BAML agents (architecture, security, synthesizer), service bindings, workflow tracking, queue verdict delivery, structured LLM extraction
ai-pitch-evaluator/	Multi-agent startup pitch evaluation with BAML	4 specialist BAML agents (market, financial, technical, verdict), service bindings, workflow tracking, queue evaluation delivery, investment decision

Multi-Runtime Examples

Example	Runtime	Description	Features
node-hello-world/	Node.js	Node.js runtime basics	Node.js APIs, CommonJS
deno-hello-world/	Deno	Deno runtime basics	ES modules, TypeScript
bun-hello-world/	Bun	Bun runtime basics	Fast startup, Web APIs
python-hello-world/	Python	Python runtime basics	WSGI/ASGI, stdlib
elixir-hello-world/	Elixir	Elixir runtime basics	OTP, Plug

Tool Integration Examples

Example	Description	Features
ffmpeg-worker/	Video processing	CLI tool binding, media processing
pdf-worker/	PDF generation	CLI tool binding, document creation

Prerequisites

1. FlameFlare running locally or deployed (e.g. on Fly.io)

2. API Token: Get your token from the setup output or re-run cd apps/server && mix run priv/repo/seeds.exs

3. FlameFlare CLI: Install the CLI to deploy workers:

   # One-time: configure GitHub Packages registry
   echo "@onixus74:registry=https://npm.pkg.github.com" >> ~/.npmrc
   echo "//npm.pkg.github.com/:_authToken=YOUR_GITHUB_PAT" >> ~/.npmrc
   
   # Install
   bun add -g @onixus74/flameflare-cli
   
   # Log in
   ff login

Environment Setup

All examples use three environment variables. Set these once before deploying or testing:

# --- Pick one ---

# Local development
export FLAMEFLARE_URL="http://localhost:4000/client/v4"

# Production (Fly.io)
export FLAMEFLARE_URL="https://flameflare.fly.dev/client/v4"

# --- Required for both ---

export FLAMEFLARE_API_KEY="your-api-token"

# Fetch your account ID automatically
export FLAMEFLARE_ACCOUNT_ID=$(curl -s "$FLAMEFLARE_URL/accounts" \
  -H "Authorization: Bearer $FLAMEFLARE_API_KEY" | jq -r '.result[0].id')

Verify the setup:

curl -s "$FLAMEFLARE_URL/user/tokens/verify" \
  -H "Authorization: Bearer $FLAMEFLARE_API_KEY"

Quick Start

Deploy any example:

cd examples/hello-world
ff deploy

The CLI reads flameflare.toml, discovers all source files, and uploads the worker. If the config includes [triggers].crons, schedules are synced automatically.

Test with curl:

# Execute a deployed worker
curl "$FLAMEFLARE_URL/accounts/$FLAMEFLARE_ACCOUNT_ID/workers/hello-world/dispatch" \
  -H "Authorization: Bearer $FLAMEFLARE_API_KEY"

Example Progression

We recommend exploring examples in this order:

1. hello-world - Basic worker structure
2. json-api - Request/response handling
3. environment-variables - Configuration management
4. router-worker - Advanced routing patterns
5. scheduled-worker - Cron triggers and scheduled tasks
6. multi-module - Multi-file ES module imports
7. wasm-module - WebAssembly binary module import
8. queue-producer-consumer - Asynchronous task processing
9. workflow-basic - Durable multi-step workflow execution
10. workflow-events - Event-driven workflows with external webhooks
11. workflow-comprehensive - Every Workflow feature in one example
12. order-processing-pipeline - Complete interconnected pipeline (Workers + Queues + Workflows + Cron + Service Bindings)

Common Patterns

Environment Setup

All examples include these standard files:

• flameflare.toml - Worker configuration (name, main module, runtime, bindings)
• src/index.js (or src/index.py, src/index.exs, etc.) - Worker entry point
• package.json - NPM configuration with deploy script
• README.md - Example-specific documentation

Runtime Configuration

Each worker can specify its runtime in flameflare.toml:

name = "my-worker"
main = "src/index.js"
compatibility_date = "2024-01-01"
runtime = "javascript"  # javascript, node, deno, bun, python, elixir

[vars]
ENVIRONMENT = "production"

Supported runtimes:

• javascript (default) - Standard Web Workers API
• node - Node.js runtime with CommonJS support
• deno - Deno runtime with TypeScript and ES modules
• bun - Bun runtime for fast startup and Web APIs
• python - Python runtime with WSGI/ASGI support
• elixir - Elixir runtime with OTP and Plug

Resource Profile Configuration

Workers can declare a resource profile to control which FLAME pool they execute on. This ensures CPU-heavy workloads get performance CPUs and more memory, while lightweight workers stay on cheaper shared infrastructure.

name = "my-video-processor"
main = "src/index.js"
resource_profile = "dedicated"

Supported profiles:

• standard (default) - Request/response handlers, API calls, compute, I/O (max_concurrency: 10, 5 min idle)
• dedicated - Heavy compute, long-running workflows, near-isolation (max_concurrency: 2, 4 hr idle)

Deployment Commands

Each example supports:

• ff deploy - Deploy to FlameFlare using the CLI

Testing

Test deployed workers with:

# Basic test
curl "$FLAMEFLARE_URL/accounts/$FLAMEFLARE_ACCOUNT_ID/workers/{worker-name}/dispatch" \
  -H "Authorization: Bearer $FLAMEFLARE_API_KEY"

# With custom path (using the /*path splat route)
curl "$FLAMEFLARE_URL/accounts/$FLAMEFLARE_ACCOUNT_ID/workers/{worker-name}/dispatch/api/users" \
  -H "Authorization: Bearer $FLAMEFLARE_API_KEY"

# With POST data
curl -X POST "$FLAMEFLARE_URL/accounts/$FLAMEFLARE_ACCOUNT_ID/workers/{worker-name}/dispatch" \
  -H "Authorization: Bearer $FLAMEFLARE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"message": "Hello!"}'

Differences from Cloudflare

These examples work on both Cloudflare and FlameFlare. FlameFlare adds multi-runtime support, resource profiles for standard/dedicated workloads, CLI tool bindings, and removes execution time limits. See the FlameFlare for Cloudflare Workers Developers guide for the full comparison.

Dashboard Monitoring

After deploying examples, visit the FlameFlare dashboard to:

• View all deployed workers
• Monitor execution logs
• Check queue message processing
• View workflow execution steps
• Monitor system metrics

Local: http://localhost:4000/dashboard Production: https://flameflare.fly.dev/dashboard

Need Help?

• Check individual example READMEs for specific instructions
• View the Getting Started guide
• Verify your API token: curl -H "Authorization: Bearer $FLAMEFLARE_API_KEY" "$FLAMEFLARE_URL/user/tokens/verify"