Metadata-Version: 2.4
Name: frugalai
Version: 0.0.1
Summary: Frugal : A minimial agentic module
Author-email: Michael John Williams <miw@slb.com>
License: MIT License
        
        Copyright (c) 2026 SLB
        
        Permission is hereby granted, free of charge, to any person obtaining a copy
        of this software and associated documentation files (the "Software"), to deal
        in the Software without restriction, including without limitation the rights
        to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
        copies of the Software, and to permit persons to whom the Software is
        furnished to do so, subject to the following conditions:
        
        The above copyright notice and this permission notice shall be included in all
        copies or substantial portions of the Software.
        
        THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
        IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
        FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
        AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
        LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
        OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
        SOFTWARE.
        
Project-URL: Repository, https://github.com/Schlumberger/frugalai
Keywords: frugal,secure,resumable,resilient,independent,hybrid
Classifier: Development Status :: 2 - Pre-Alpha
Classifier: Environment :: Other Environment
Classifier: Intended Audience :: Developers
Classifier: Intended Audience :: Science/Research
Classifier: License :: Other/Proprietary License
Classifier: Natural Language :: English
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3.12
Classifier: Topic :: Scientific/Engineering
Classifier: Topic :: Scientific/Engineering :: Physics
Classifier: Topic :: Software Development :: Libraries
Classifier: Topic :: Utilities
Requires-Python: >=3.12
Description-Content-Type: text/markdown
License-File: LICENSE
Dynamic: license-file

# frugalai

A minimal python module for AI-enabled workflow automation, which encourages cost discipline efficient model use through repeatable routing and lightweight abstractions for building reliable, scalable calculation graphs.

Whether automating the boring repetitive tasks, or dreaming of 10x productivity gains, it is hard to ignore the impact of the LLM-based agentic workflow.

Beneath the technology stack there are some relatively simple concepts that we believe are worth learning. Armed with that understanding, whether you a creating
workflows for yourself as a hobby, or managing a team building the AI-first business that will change the world, you can start to demand some requirements are met:
- **Frugal** – minimize unnecessary token use and have visibility over projected costs
- **Secure** – no direct path from generated text to commandline execution
- **Resumable** – be able to restart from any point in the agentic computation graph & recover from lost connection to the LLM
- **Resilient** – retain "stick-and-rudder" skills of your subject matter experts (SMEs)
- **Independent** – right-sizing the choice of LLM for every task
- **Hybrid** - non-LLM-based agents interacting directly with LLM-based agents.

The **frugal** module is designed so that you can build and deploy without the stack. The code is designed to be small and illustrative. Each file is small enough that you could paste it into your favourite LLM assistant and ask for it be explained. 

The **thrifty** example application offers no-code development of workflows, designed around multiple system prompts. It is a human-in-the-loop version of the automated "learning" of full enterprise-grade agent. YOu can use it to learn how actor-critic approaches operate; or to build and deploy complete AI-enabled workflows. 

The **examples** show what we have built, and we welcome contributions! We also welcome contributors to the core **frugal** library, noting that the priority is to keep everything small enough to read and learn from.



