Please visit our website!

Welcome to the LSP Client ecosystem — a professional-grade toolkit and agent-native ecosystem designed to transform the Language Server Protocol (LSP) into actionable repository intelligence for AI Agents and developers.

Our ecosystem is built as a layered stack, progressively evolving raw language server capabilities into specialized reasoning "skills" for coding agents.

The industrial-strength foundation. A modern, async-first Python SDK for fine-grained LSP control.

• Protocol Native: Full support for LSP specifications with automatic capability negotiation.
• Robustness: Type-safe, async-first (AnyIO), and production-ready.
• Container Registry: Pre-built, isolated environments for 20+ language servers, enabling instant, zero-config deployment.

Language Server Agent Protocol. A semantic abstraction layer that transforms granular LSP operations into agent-friendly snapshots.

• Agent-Centric: Converts raw JSON-RPC data into structured Markdown/JSON optimized for LLM consumption.
• Semantic Context: Provides high-signal context (e.g., "Summarize this class" instead of just "Get symbols").

A universal command-line interface that brings LSP and LSAP capabilities to any terminal or script.

• Human & Machine Friendly: Use it for manual exploration or as a backend for automation scripts.
• Universal Interface: One CLI to rule all language servers.

High-level repository analysis capabilities designed specifically for Coding Agents (like Opencode).

• Deep Understanding: Provides cross-file reference analysis, project-wide symbol mapping, and architectural insights.
• Plugin for Agents: Acts as a skill that agents can invoke to perform complex repository-wide tasks via LSP.

We provide optimized, pre-built container images for instant, zero-config deployment:

Integrate language intelligence into your Python projects:

import anyio
from lsp_client.clients.pyright import PyrightClient, PyrightContainerServer
from lsp_client.common import Position

async def main():
    # Run Pyright in an isolated container
    async with PyrightClient(server=PyrightContainerServer()) as client:
        definitions = await client.request_definition_locations(
            file_path="app.py",
            position=Position(line=10, character=5)
        )
        for loc in definitions:
            print(f"Found definition: {loc.uri} at {loc.range}")

anyio.run(main)

We are building the future of AI-native software engineering.

• Contribute: Add new servers to our Container Registry.
• Extend: Help us evolve LSAP for better agent reasoning.
• Integrate: Use LSP Skills to empower your own AI coding assistants.

Built with ❤️ by the LSP Client team.