Supercharge AI Agents, Safely
n2-QLN: Intelligent Tool Router & Semantic Search Layer for MCP. Connect 1,000+ tools through 1 interface. Prevent AI model confusion and maximize context window efficiency
local-first MCP discovery runtime for agents — search before invoke, reduce prompt bloat, and route to local MCP tools
Semantic routing for MCP tools - .NET library that indexes MCP tool definitions and returns the most relevant tools via vector search
A retrieval-gated skill architecture for LLM agents that scales to hundreds of tools by exposing only the top-K relevant capabilities per request.
A governed skill system for long-horizon human-AI collaboration.
Orchestration MCP Server - Routes tool calls to MCP backends
Budget-aware context compilation and context firewall for tool-heavy AI agents.
Resource-first MCP hub for routing upstream MCP servers through one small local tool surface.
Intelligent tool routing extension for pi coding agent - automatically selects the best tool stack for each task
Intelligent routing layer for AI agents: recommends the best MCP server and LLM for any task, scored on 132+ real benchmark executions.
LangGraph is a powerful framework built on LangChain that enables the creation of stateful, multi-step, and agentic workflows using directed graphs. It simplifies complex LLM orchestration by allowing conditional branching, memory, and tool integrations in a visual and modular way.
Smart MCP router that routes tools by intent and reduces context rot for LLM hosts
Route semantic tool requests through one MCP router to simplify AI access to 1,000+ tools in n2-qln
An AI assistant with hot-pluggable MCP tool servers — one Router Agent, dynamic tool discovery, and LLM-powered routing. Built with FastAPI, Redis, Ollama, Streamlit, and Prometheus.
On-device tool-routing models for the kAir iOS super-app. Qwen3.5-0.8B + LoRA on Apple Silicon (MLX-LM); 5-tool closed catalog with frozen JSON schema; three-layer artifact pipeline (HF → GGUF → Core ML / ONNX). Research / experimental.
The pre-LLM reflex layer for AI agents. library with a live training server, microsecond latency, learns from LLM corrections — System 1 to your LLM's System 2.
Offline-first FastMCP/MCP tool routing kit for OpenAI-compatible gateway planning
Pick the right tools for an agent task. Boring baseline, open dataset (14K traces from public agent benchmarks).
Reduce MCP tool overload with a single Go binary that indexes tools locally and serves search-before-invoke via SQLite
Add a description, image, and links to the tool-routing topic page so that developers can more easily learn about it.
To associate your repository with the tool-routing topic, visit your repo's landing page and select "manage topics."