GitHub Copilot CLI can now use Language Server Protocol servers for real code intelligence, replacing text-based heuristics with precise, structured answers about your code. Announced June 10, 2026, the LSP Setup skill lets the command-line agent resolve types across dependencies, jump to definitions in third-party libraries, and find symbol references across 14 languages including Java, Python, TypeScript, and Rust.
How to Give Copilot CLI Code Intelligence
The setup is a downloadable skill, not a new install. Grab the LSP Setup skill from the Awesome Copilot project and add it to your local Copilot CLI configuration. Once active, the agent talks to a real language server instead of grepping your files, so when you ask about a type or function it returns the resolved definition rather than a best-guess from pattern matching. That removes the workarounds the agent used to rely on, like extracting JAR files or crawling node_modules to understand dependencies.
Why It Matters for Creators
For anyone building tools, plugins, or sites with an AI coding agent, the quality of answers depends on how well the agent understands the codebase. The Language Server Protocol is the same technology that powers go-to-definition and type hints in editors like VS Code, so wiring it into the CLI agent gives it the same grounding a human developer gets. Fewer hallucinated APIs and fewer wrong guesses mean less time spent correcting the agent. This follows the broader move toward agent skills in coding tools that extend what CLI and IDE agents can do.
Key Details
Languages: 14 supported, including Java, Python, TypeScript, and Rust.
Capability: Resolve types across dependencies, navigate to definitions in third-party libraries, find references, and pull documentation through the language server.
Availability: Live now as a downloadable skill from the Awesome Copilot catalog for the local Copilot CLI.
What to Do Next
If you already run Copilot CLI, install the LSP Setup skill and test it on a project with heavy third-party dependencies, where type resolution makes the biggest difference. Compare an answer before and after enabling the skill to gauge the accuracy lift for your stack before standardizing it across your projects.
