Syntic Code

Agentic Coding for the Terminal, the SDK, and the Cloud Desktop

Syntic Code is an agentic coding platform built in Rust and Bun. It gives engineering teams a fast terminal CLI, programmable SDK, and Cloud Desktop — all running on the model and infrastructure of your choice, with enterprise controls baked in.

What You Get

Surface	Description
`syntic` CLI	Native single-binary CLI for macOS, Linux, and Windows. Compiled with Bun for sub-100ms startup.
Syntic SDK	TypeScript, Python, and Rust client libraries for building your own coding agents on Syntic’s runtime.
Cloud Desktop	Browser-based coding environment with full agent capabilities — no local install required.
Skills System	Composable agent behaviors. Install from the Syntic Marketplace or author your own.
MCP Native	First-class support for Model Context Protocol servers — connect any tool, any data source.
Sub-Agents	Dispatch parallel specialized agents for codebase exploration, planning, review, and execution.

Why Syntic Code

	Syntic Code	Other Coding Agents
Model choice	Any model — Kimi, GPT, Claude, Gemini, local	Locked to one vendor
Deployment	Self-hosted, cloud, or hybrid	Cloud-only
Compliance	SOC 2, HIPAA-ready, SSO, audit log	Limited
White-label	Full re-skin for resellers and enterprises	No
Migration	Drop-in compatible config (`.syntic/` or `.claude/`)	Lock-in

Quick Install

macOS / Linux

curl -fsSL https://syntic.ai/install.sh | sh

Windows (PowerShell)

iwr -useb https://syntic.ai/install.ps1 | iex

Verify

syntic --version

Then run syntic in any project folder to start an interactive session.

Documentation Map

Get Running

Quickstart — Install, sign in, run your first task
Installation — Per-platform install options, system requirements
Cloud Desktop — Browser-based coding environment

Core Concepts

CLI Reference — Every command, flag, and environment variable
SDK Reference — TypeScript, Python, Rust client libraries
Skills — Composable agent behaviors and the Skills Marketplace
Hooks — Lifecycle hooks for automation and policy
MCP Servers — Integrate external tools via Model Context Protocol
Sub-Agents — Parallel specialized agents for complex workflows
Plan Mode — Approve plans before execution
Slash Commands — Built-in and custom slash commands

Configuration

Settings — settings.json schema and precedence
Permissions & Security — Tool approval, sandboxing, allowlists
Models — Configure providers, switch models, set fallbacks
IDE Integrations — VS Code, JetBrains, Vim, Zed

Enterprise & Migration

Enterprise — SSO, audit, compliance, self-hosted deployments
Migrate from Claude Code — Drop-in replacement guide

Reference

Troubleshooting — Common errors and fixes
Changelog — Release notes by version