How is the Zed setup different from Visual Studio Code-based editors?

Zed uses `.zed/settings.json` for editor behavior and a `.rules` file for repo guidance, so the setup is intentionally different from VS Code-based editors.

Why does Zed use append mode for the rules file?

Append mode means Ultracite can add its rules to the existing `.rules` file without replacing everything that is already there, which is helpful for teams layering repo guidance over time.

Who is the Zed setup best for?

Teams that care about editor responsiveness, native tooling, or collaborative editing tend to benefit most from keeping Zed's native config model intact while still standardizing on Ultracite.

Ultracite for Zed

Zed uses a native Ultracite setup with `.zed/settings.json` for editor behavior and an appendable `.rules` file for repo guidance, making it a strong fit for teams that want speed without config drift.

Setup files for Zed

Ultracite uses these Zed settings, rules, and hooks files to keep editor behavior and AI output aligned with your repo standards.

Workspace settings

This native settings file keeps Zed's formatter, code actions, and language tooling aligned with Ultracite.

.zed/settings.json

JSON config

{  "format_on_save": "on",  "formatter": "language_server",  "lsp": {    "typescript-language-server": {      "settings": {        "typescript": {          "preferences": {            "includePackageJsonAutoImports": "on"          }        }      }    }  },  "languages": {    "JavaScript": {      "code_actions_on_format": {        "source.fixAll.biome": true,        "source.organizeImports.biome": true      },      "formatter": {        "language_server": {          "name": "biome"        }      }    },    "TSX": {      "code_actions_on_format": {        "source.fixAll.biome": true,        "source.organizeImports.biome": true      },      "formatter": {        "language_server": {          "name": "biome"        }      }    },    "TypeScript": {      "code_actions_on_format": {        "source.fixAll.biome": true,        "source.organizeImports.biome": true      },      "formatter": {        "language_server": {          "name": "biome"        }      }    }  }}

Why Zed teams choose Ultracite

A high-performance, multiplayer code editor built in Rust with built-in AI assistance.

Native configuration for a native editor

Zed uses its own native settings file, so Ultracite can tailor the setup to Zed instead of pretending every editor is a VS Code clone.

Append-mode rules fit iterative teams

Ultracite's `.rules` file works well with Zed's append-mode workflow, letting you keep existing instructions and add standards incrementally.

Keeps Zed's performance-focused workflow intact

The setup supports fast local editing, multiplayer sessions, and AI assistance without forcing a separate VS Code compatibility layer.

Workflow tips for Zed

Keep Ultracite present in the day-to-day Zed workflow with a few editor-specific habits, not just a one-time setup.

Generate .zed/settings.json so Zed handles formatting, code actions, and TypeScript behavior the way Ultracite expects.

Append Ultracite guidance into .rules so Zed keeps any existing repo instructions and gains a shared coding standard layer.

Use the native Zed files as the source of truth instead of trying to mirror a VS Code setup in a different editor model.

Zed FAQ

Editor-specific answers for teams rolling out Ultracite in Zed.

Compare Zed with other Ultracite-supported editors

These nearby setups make it easier to compare how Ultracite handles shared settings, AI rules, and editor-specific workflow details.

IBM BobIBM's AI coding assistant

IBM Bob works well with Ultracite when you pair shared Visual Studio Code workspace settings with `AGENTS.md` rules—matching the standard format while keeping editor behavior predictable.

KiroAWS's spec-driven IDE

Kiro pairs well with Ultracite when you combine shared VS Code workspace settings with a dedicated steering file that guides spec-driven AI work before code is generated.

TraeByteDance's AI IDE

Trae works well with Ultracite when you keep the standard VS Code workspace settings in place and add a dedicated project rules file that teaches the AI how your repo expects code to look.