#llm

PDBe MCP Servers integrate Protein Data Bank Europe resources with LLMs via Model Context Protocol. Provides seamless access to protein structure data through API tools and graph database schema assistance for intelligent Cypher query generation, bridging structural biology and AI research.

Local execution trees for TypeScript AI agents. agent-inspect helps you understand what happened inside an AI agent run — locally. It turns manual steps, tool calls, LLM calls, structured logs, failures, durations, and run metadata into readable execution trees you can inspect from the terminal. It is built for TypeScript/Node.js developers..

The developper starter pack for document processing

TUI and CLI for browsing AI models, benchmarks, coding agents, and statuses for AI providers.

AI-powered security alert management that reduces noise and accelerates response time

Python framework for composable LLM workflows.

Brigade — Your personal intelligence, built enterprise-grade

🥤 RAGLite is a Python toolkit for Retrieval-Augmented Generation (RAG) with DuckDB or PostgreSQL

PHP library for building agentic applications.

Symfony integration bundle for Model Context Protocol (via official mcp/sdk)

Spine — governed, provenance-grounded autonomous software delivery (installs as agent-orchestrator).

Aiden — Autonomous AI agent that operates your computer with prompts: browser control, terminal execution, workflows, tools, recovery systems, and persistent memory. Built solo. AGPL-3.0.

Your Intelligence, Orchestrated. Every builder. Every team. Every agent. For Everyone.

C# SDK for the Cohere API -- chat completions, embeddings, and reranking

A composable framework for building reliable LLM agents on a harness you fully control. The same code runs the agentic loop on every model tier, local 4B to frontier with tool-call healing, verification, and durable crash-resume, on a typed 12-phase engine where every step emits an event you can hook into. Effect-TS, MCP-native, MIT

Whale — blazingly fast, terminal-first AI coding agent for DeepSeek. ~98% prompt cache hit rate, 1M context, MCP tools, dynamic workflows.

Define AI agent roles in YAML and run them anywhere: CLI, API server, or autonomous daemon

Turn your company's scattered knowledge into AI ready Books ✨

The agent loop for Rust — stream from 7 LLM protocols, run tools, loop until done.

llama_cpp.rb provides Ruby bindings for llama.cpp

One config for every AI coding agent — cross-tool config + a cognitive substrate (memory, blast-radius, guardrails) for Claude Code, Codex, Cursor, Gemini, Aider, and more.

A full featured, enterprise grade rust MCP SDK

Dulus Ai — Free Agentic AI, Making Gemini web cappable of running bash commands in your terminal! [Gui, Web, Cli, Telegram, 2,000 MCP, 100K Skills . LiteLLM (100+ providers), local models via Ollama, /lang in 34 languages, Mesa Redonda, I create the first utility coin that can be used 100% as AI quota or Fuel, is called $Dulus

An open-source, extensible, self-hosted agent workspace with multi-runtime support for Codex, Claude Code, and more, plus reusable local apps for custom interfaces and services.

Graph-based tool retrieval for LLM agents — 248 tools → 82% accuracy, 79% fewer tokens. Zero dependencies. OpenAPI / MCP / LangChain.

Let AI agents query and update Postgres/MySQL without raw SQL, unrestricted schema access, or database credentials in the model; writes stay reviewed.

Production focused Self-harnessed LM runtime (RLM) that allows the LM to call its sub-lm with DSPy signatures. Define your inputs, outputs, and tools — the model handles its own control flow. Get fully interpretable trajectories and performance that scales directly with model improvements. Without context rot.

The only fully local production-grade Super SDK that provides a simple, unified, and powerful interface for calling more than 200+ LLMs.

A secure JavaScript sandbox designed for safe AI agent code execution. Protects against code injection, prototype pollution, and sandbox escapes. The security layer that makes AI-generated code safe to run.

Consumer-aware compression for embedding indexes and LLM KV caches — compress by the metric the downstream consumer actually uses. PCA-Matryoshka + TurboQuant (27x @ 99.8% recall@10), asymmetric K/V, CUDA/Triton kernels, vLLM plugin, replayable CI-gated claims. MIT.