griptape

griptape is a modular Python framework for LLM workflows, tools, memory, and data that enables developers to:

1. 🚰 Build sequential LLM pipelines and sprawling DAG workflows for complex use cases.
2. 🧰️ Augment LLMs with chain of thought capabilities and integrate external tools, such as calculators, web search, spreadsheet editors, and API connectors.
3. 💾 Add memory to AI pipelines for context preservation and summarization.

Main Packages

griptape-flow

Build LLM workflows and pipelines with memory, rules, and chain of thought reasoning.

---

griptape-core

Integrate other services and functionality into LLMs as tools; run tools in any environment (local, containerized, cloud, etc.); convert tools into underlying middleware abstractions, such as ChatGPT Plugins, LangChain tools, and Fixie.ai agents.

---

griptape-tools

Official Griptape tools registry.

Documentation

Please refer to Griptape Docs for:

• Getting started guides.
• Core concepts and design overviews.
• Examples.
• Contribution guidelines.

Quick Start

First, install griptape that includes all core modules:

pip install griptape -U

Second, configure an OpenAI client by getting an API key and adding it to your environment as OPENAI_API_KEY. griptape uses OpenAI Completions API to execute LLM prompts and to work with LlamaIndex data structures.

With griptape, you can create structures, such as Pipelines and Workflows, that are composed of different types of steps. You can also define structures as JSON objects and load them into griptape dynamically. Let's define a simple two-step pipeline that uses tools:

from decouple import config
from griptape.tools import WebScraper, Calculator
from griptape.flow import utils
from griptape.flow.memory import PipelineMemory
from griptape.flow.steps import PromptStep, ToolkitStep
from griptape.flow.structures import Pipeline
from griptape.flow.utils import ToolLoader


scraper = WebScraper(
    openai_api_key=config("OPENAI_API_KEY")
)
calculator = Calculator()

pipeline = Pipeline(
    memory=PipelineMemory(),
    tool_loader=ToolLoader(
        tools=[calculator, scraper]
    )
)

pipeline.add_steps(
    ToolkitStep(
        tool_names=[calculator.name, scraper.name]
    ),
    PromptStep(
        "Say the following like a pirate: {{ input }}"
    )
)

pipeline.run("Give me a summary of https://en.wikipedia.org/wiki/Large_language_model")

print(utils.Conversation(pipeline.memory).to_string())

Boom! Our first multistep LLM pipeline with memory that used tools generated a conversation:

Q: Give me a summary of https://en.wikipedia.org/wiki/Large_language_model
A: Arr, me hearties! Large language models have been developed and set sail since 2018, includin' BERT, GPT-2, GPT-3 [...]

Versioning

griptape is in early development and its APIs and documentation are subject to change. Until we stabilize the API and release version 1.0.0, we will use minor versions (i.e., x.Y.z) to introduce breaking features and patch versions (i.e., x.y.Z) for bug fixes.

Contributing

Contributions in the form of bug reports, feature ideas, or pull requests are super welcome! Take a look at the current issues and if you'd like to help please submit a pull request with some tests.

License

griptape is available under the Apache 2.0 License.