Skip to main content

Crimson Langgraph Reflection

A modular implementation of LangChain's Reflexion example, designed to make agent development more accessible, extensible, and reusable.

Overview

This project transforms the original LangChain Reflexion tutorial into a fully modular package. The Reflexion pattern enables LLM agents to:

  1. Answer questions
  2. Reflect on and critique their own answers
  3. Generate search queries to gather more information
  4. Revise answers based on new information

By modularizing this pattern, this package aims to:

  • Make Reflexion-style agents easier to implement in your projects
  • Provide a solid foundation for further customization and extension
  • Encourage reusability of core Reflexion components

Installation

pip install crimson-langgraph-reflection

Key Components

The package is organized into several main modules:

  • graph.py: Defines the LangGraph state machine that orchestrates the reflection process
  • responder/: Contains components for generating answers, reflections, and search queries
    • base.py: Core classes for response generation and validation
    • prebuilt.py: Ready-to-use responder implementations
  • tool.py: Search tool integration for information gathering
  • ui.py: Simple interface for running the Reflexion agent
  • llm.py: LLM configuration (defaulting to Claude 3.5 Sonnet)

Quick Start

from crimson.langchain_reflexion.ui import stream_shortcut

# Run with default settings (5 iterations)
stream_shortcut(question="What is LangChain?")

# Or customize the number of iterations
stream_shortcut(
    question="What is LangChain?",
    max_iterations=3
)

# Use a custom LLM
from langchain_anthropic import ChatAnthropic
custom_llm = ChatAnthropic(model="claude-3-opus-20240229")

stream_shortcut(
    question="What is LangChain?",
    llm=custom_llm
)

Examples

For complete examples, check out:

Extension and Customization

This package is designed to be extensible. You can:

  1. Create custom responder implementations by extending the base classes
  2. Modify the graph structure to add additional steps
  3. Integrate different search or reasoning tools
  4. Adjust the system prompts and reflection criteria

Requirements

  • Python ≥ 3.9
  • LangChain
  • LangGraph
  • Anthropic API access (for default LLM)

License

MIT

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Acknowledgements

This project is based on LangChain's Reflexion tutorial. Special thanks to the LangChain team for their innovative work on LLM agent frameworks.

Metadata

Release files for crimson-langgraph-reflection 0.1.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for crimson-langgraph-reflection 0.1.0
File Size Uploaded Provenance Actions
crimson_langgraph_reflection-0.1.0.tar.gz 9.7 kB Without provenance Details for crimson_langgraph_reflection-0.1.0.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for crimson-langgraph-reflection 0.1.0
File Interpreter ABI Platform Actions
crimson_langgraph_reflection-0.1.0-py3-none-any.whl Python 3 none any Details for crimson_langgraph_reflection-0.1.0-py3-none-any.whl

Total release size: 20.4 kB

Release files / crimson_langgraph_reflection-0.1.0.tar.gz

Download URL crimson_langgraph_reflection-0.1.0.tar.gz
Size 9.7 kB
Tags Source
SHA-256 checksum
How to use checksums
6cf56e9c963af578b8dd441388944dafdbc728859784bf7e87019343f21e8ad8
BLAKE2b-256 checksum
How to use checksums
44b2c1a8e81bfc47bda6a0cf45545435640fa7be4b4686356d475af5ad3ad658
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.1.0 CPython/3.9.21

Release files / crimson_langgraph_reflection-0.1.0-py3-none-any.whl

Download URL crimson_langgraph_reflection-0.1.0-py3-none-any.whl
Size 10.7 kB
Tags Python 3
SHA-256 checksum
How to use checksums
0070f66d90f92db25923014ce6828a3311d6ce3fe0226c0c7521f68c74ed8b71
BLAKE2b-256 checksum
How to use checksums
6b7b15af81551de0522a2019218dcbfc9578f5cce770e43080be8b5ddd673d27
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.1.0 CPython/3.9.21

Security for crimson-langgraph-reflection 0.1.0

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Release history Release notifications | RSS feed

This release

0.1.0 This release

2 release files No provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page