Tenets Add to favorites

Upvote

Last update time : 2025-12-12 11:26:27

Tenets is a free, open-source MCP server and Python library enhancing AI coding assistants. It uses BM25, TF-IDF, and semantic embeddings to deliver highly relevant, privacy-preserving code context.

Tenets is a powerful, free, and open-source solution designed to dramatically improve the performance and relevance of AI coding assistants. It serves as an MCP server, Python library, and command-line interface (CLI) dedicated to building highly intelligent and Large Language Model (LLM)-ready code context.

The core strength of Tenets lies in its sophisticated, multi-factor relevance scoring engine. By combining traditional Natural Language Processing (NLP) techniques like **BM25** and **TF-IDF** with project import graphs and optional semantic embeddings, the system is able to meticulously analyze, rank, and distill the most critical and relevant files from a codebase. This process ensures that AI assistants are fed only the highest-signal context, eliminating noise and improving output quality.

Furthermore, Tenets introduces a unique feature: the automatic injection of project 'tenets' or guiding principles directly into the prompt. This mechanism helps enforce consistent, AI-driven code guidance across the entire team. It also provides robust session and stateful context management, along with automated context summaries, which significantly reduces 'context drift'—a common issue where the model forgets earlier details of a conversation.

Crucially, the entire system operates **100% locally**, ensuring absolute privacy and data security. Tenets offers native MCP support for popular AI coding environments such as **Cursor**, **Claude**, and **Windsurf**. By providing high-signal, privacy-preserving context, Tenets not only speeds up the onboarding process for new developers but also guarantees that the AI's assistance is consistently relevant and aligned with project standards.

Pricing : Open Source

Web Address : Tenets

Tags : AI coding assistant open-source LLM context Tenets MCP server Python library BM25 TF-IDF semantic embeddings code relevance privacy-preserving context management