llm-wrapper

Agent SDK for Go - with Temporal & Restate support for durable execution, multi-agent support, and OTeL tracing support

A robust CLI powering autonomous web agents. Seamlessly integrate Playwright, browser-use, and LangChain to automate your daily web workflows.

Python wrapper that simplifies the Google Gemini API

🐍 Script to compare llm performances with statistics

PolyChat is a focused workspace for thinking, creating, and working with AI.

— or your Universal MCP Hub (Sandboxed) — or a secure All-In-One AI wrapper with dual interface: REST + MCP aka: a clean, secure starting point for your own projects. Pick the description that fits your use case. They're all correct.

From 90's cartoon, Courage the Cowardly dog, A LLM Wrapper and TTS based off of courage's snarky computer in the attic. The Bagge family's Computer AI has finally been brought into real life.

🤖 Minimal CLI agent using single-turn ReAct pattern with OpenAI LLM wrapper. Intelligently selects between calculator, string operations, and direct responses. Extensible tool system with comprehensive testing.

Modified version of Le Wagon Data Science capstone project (March 2024)

Privacy-preserving LLM wrapper that anonymizes PII locally before it reaches a cloud LLM and restores it in the response. Drop-in OpenAI replacement, multi-SDK support, CLI, REST API, and audit logging. No Docker required.

An ultra-lightweight, asynchronous Python 3.14 client to centralize and unify calls across AI providers (Cerebras, Cohere, DeepSeek, Gemini, Groq, Mistral, Nvidia NIM, OpenRouter, OrcaRouter). Powered by uv.

LLM based git-hook script that leverages the model to generate conventional commit style message based on context gather around the diff. In Case it doesn't work for you, create an Issue. Works on my machine LOL

Automated quiz app for familiarising with a code repository

Unified LLM abstraction layer with fallback routing, retry policies, and token/latency telemetry.