# pocketflow-throttled

> Rate-limited parallel execution patterns for LLM APIs and web services. Use when you need to prevent 429 errors and optimize throughput.

- Author: Jason-AI-lab
- Repository: Jason-AI-lab/pocketflow-throttled
- Version: 20260203132906
- Stars: 1
- Forks: 0
- Last Updated: 2026-02-06
- Source: https://github.com/Jason-AI-lab/pocketflow-throttled
- Web: https://mule.run/skillshub/@@Jason-AI-lab/pocketflow-throttled~pocketflow-throttled:20260203132906

---

---
name: "pocketflow-throttled"
description: "Rate-limited parallel execution patterns for LLM APIs and web services. Use when you need to prevent 429 errors and optimize throughput."
---

# PocketFlow Throttled

Reusable patterns and skills for working with `pocketflow-throttled`, a library for rate-limited parallel execution of LLM APIs and web services.

## Quick Decision Guide

**Choose the right pattern based on your needs:**

- **Getting 429 errors from LLM API?** → [01-basic-node-throttling.md](01-basic-node-throttling.md)
- **Don't know exact rate limits?** → [02-adaptive-throttling.md](02-adaptive-throttling.md)
- **Each item needs multiple API calls?** → [03-flow-level-throttling.md](03-flow-level-throttling.md)
- **Multiple services calling same API?** → [04-shared-limiters.md](04-shared-limiters.md)
- **Different users need different configs?** → [05-parameterized-flows.md](05-parameterized-flows.md)
- **Using OpenAI/Anthropic/Google?** → [06-using-presets.md](06-using-presets.md)
- **Complex pipeline with many API calls?** → [07-nested-throttling.md](07-nested-throttling.md)

## Available Patterns

### Core Skills

1. **[Basic Node Throttling](01-basic-node-throttling.md)** - Fixed rate limits for single-node batch processing
2. **[Adaptive Throttling](02-adaptive-throttling.md)** - Self-tuning concurrency that adapts to API responses
3. **[Flow-Level Throttling](03-flow-level-throttling.md)** - Controlling concurrent flow instances
4. **[Shared Limiters](04-shared-limiters.md)** - Global rate limiter coordination across components
5. **[Parameterized Flows](05-parameterized-flows.md)** - Passing different parameters to each flow instance
6. **[Using Presets](06-using-presets.md)** - Pre-configured settings for popular APIs
7. **[Nested Throttling](07-nested-throttling.md)** - Combining flow and node-level rate limits
8. **[Architecture Patterns](08-architecture-patterns.md)** - Core design patterns and flow mechanics

## When to Use This Skill

Use `pocketflow-throttled` patterns when you need to:
- Process batches of items through LLM APIs without hitting rate limits
- Maximize throughput while staying within API quotas
- Handle unknown or variable rate limits
- Coordinate rate limiting across multiple services
- Build complex multi-step pipelines with rate-limited APIs

## Complexity Levels

- **Beginner**: Skills 01, 06
- **Intermediate**: Skills 02, 03, 05
- **Advanced**: Skills 04, 07, 08

## Learning Path

1. Start with **Skill 01** - Understand basic throttling
2. Read **Skill 08** - Learn the architecture
3. Try **Skill 02** - Add adaptive behavior
4. Explore **Skill 03** - Scale to flow-level
5. Master **Skills 04, 05, 07** - Advanced patterns

## Working Examples

See [examples/](examples/) directory for complete, runnable code examples:

- **[01-basic-node-throttling.py](examples/01-basic-node-throttling.py)** - Fixed rate limits demo
- **[02-adaptive-throttling.py](examples/02-adaptive-throttling.py)** - Self-tuning concurrency demo
- **[03-flow-level-throttling.py](examples/03-flow-level-throttling.py)** - Multi-node pipeline demo
- **[04-shared-limiters.py](examples/04-shared-limiters.py)** - Global limiter coordination demo
- **[05-parameterized-flows.py](examples/05-parameterized-flows.py)** - Different configs per item demo
- **[06-using-presets.py](examples/06-using-presets.py)** - API preset configurations demo

All examples use a mock LLM API simulator, so you can run them without real API costs. See [examples/README.md](examples/README.md) for setup instructions.

## Additional Resources

- **Source Code**: `/cookbook/rate_limited_llm_batch/`
- **Architecture Docs**: `/docs/`
- **Main Documentation**: `/CLAUDE.md`