Docs
Everything LingCode, in one map.
Start with core concepts, move to AI workflows, build and run, ship, then deep reference. Jump to the section you need.
Start here · Core concepts · AI workflows · Build & run · Ship · Deploy fundamentals · Reference
Start here
Install, first launch, first AI session, first deploy. ~10 minutes end-to-end.
Getting Started
Install LingCode, add your first API key, open a project, try a slash command, run a build. The 10-minute onboarding.
Development Setup
Backend, database, frontend. Postgres/MySQL/Redis locally, migrations, .env patterns, five stacks end-to-end.
API keys & providers
Set up Anthropic, OpenAI, Gemini, DeepSeek, Qwen, or local Ollama. Keys live in Keychain.
Core concepts
The four things that make LingCode structurally different from Cursor, Copilot, and other AI editors. Understand these and the rest of the docs click into place.
Work protection
Pre-edit snapshots of project.pbxproj, Info.plist, Gradle, entitlements. Per-hunk diff review. Semantic time-travel undo. Worktree isolation. Nothing the agent changes is one-way.
IDE-aware agent
The agent calls BuildLogService, LLDB, SimulatorController as in-process Swift APIs — not shelled-out commands. It reads build state, debugger, sim logs live.
Your keys. No proxy.
Direct-to-provider architecture. Your API keys live in the macOS Keychain. Prompts go from your Mac to Anthropic/OpenAI/Gemini with zero LingCode servers in between.
Local models & offline
One-click Ollama integration. Run Qwen 2.5 Coder, Llama 3, Mistral locally. Full IDE functionality with zero network traffic. Cursor cannot do this.
AI workflows
How to actually get leverage from the agent day-to-day. Patterns, habits, and tool integrations.
AI development best practices
Ten habits for Swift/iOS developers using LingCode with Claude: prompting, code review, Git, context, work protection, and IDE-aware prompting.
Skills & slash commands
25+ built-in AI workflows — /review, /explain, /refactor, /fix, /commit, /security, /mode, and more. Plus custom skills.
Multiple Claude sessions
Run parallel Claude Code tabs on the same project. Keep feature work and refactoring in separate contexts.
IDE-aware prompts
Prompt patterns that only work when the agent can read your live build log, debugger, and sim logs. "Why did this build fail?" without pasting errors.
Work protection workflow
How to check the recovery panel, use worktree isolation for risky experiments, and trust semantic undo over git reset.
Remote control from iPhone/iPad
SSH + Tailscale + tmux. Drive Claude Code on your Mac from anywhere with full chat control.
Build & run
The toolchain side: iOS Simulator, Android Emulator, debuggers, signing. LingCode's native integrations with Apple and Google SDKs.
Apple toolchain (Xcode-side)
xcodebuild, iOS Simulator via simctl, LLDB debugger, code signing from Keychain, scheme editor, SPM packages, archive & TestFlight.
Android toolchain
Gradle builds, AVD emulators via emulator & adb, logcat streaming, Kotlin debugging via JDWP/DAP, APK/AAB output.
Editor & LSP
SourceKit-LSP for Swift, Kotlin LSP, TypeScript, Python, Go, Rust. Tree-sitter highlighting. Per-hunk AI diff review.
Database tools
PostgreSQL workspace (connection manager, query runner with autocomplete, row-level edit, EXPLAIN trees). MySQL inspector. SQLite inspector & schema editor. Agent can read your live schema.
Codebase index
Local, persistent symbol index. Tree-sitter + swift-syntax. Fast lookup, jump-to-symbol, agent context. Zero cloud upload.
Ship
Every deploy target, with the exact one-time platform setup each needs. Magic Deploy detects your project and runs the right flow.
App Store (iOS + macOS)
Archive, export, and upload via App Store Connect API key. Platform auto-detected from Xcode scheme.
Google Play (Android)
Signed AAB via Play Developer API. Remote preflight verifies access before gradle runs.
Web platforms
Vercel, Netlify, Railway, Fly.io, Heroku. CLI install + auth per platform.
Your own server
Before-launch & After-deploy script cookbook for VPS / droplet / bare-metal rsync deploys.
Deploy hub
All deploy targets on one page, with "which should I pick" guidance.
Deploy FAQ
Real questions devs ask when setting up deploy. App Store quirks, Play version codes, CLI auth.
Deploy fundamentals
Not required to ship. Essential when things break. Written plain-language — no "read the Apple docs" hand-waving.
How App Store code signing works
The trust chain from Apple's root CA down to the signature on your binary. Why provisioning profiles exist.
JWT auth in 5 minutes
Why every deploy tool uses JSON Web Tokens — and why they beat API keys.
Why versionCode breaks your second deploy
Build numbers, monotonic integers, and the rules each store enforces.
What a service account really is
Google's identity for machines, and why it's strictly better than an API key.
Annotated file examples
Side-by-side views of .p8, service-account JSON, and ExportOptions.plist with every field explained.
Glossary
Every term from the deploy guides in plain English. 28 entries.
Reference
FAQ, competitive comparisons, support, and community.
Support & FAQ
26 detailed Q&As on privacy, work protection, IDE awareness, local models, troubleshooting. Plus bug reports and direct contact.
Why not Cursor?
Five structural differences: trust, IDE-awareness, no proxy, local models, and iOS tooling. Plus the Electron ceiling.
Full feature list
Every feature LingCode ships, grouped by area. The exhaustive reference.
Skills reference
The 25+ built-in slash commands (/review, /commit, /explain, /batch) with full usage docs and examples.
lingcode CLI
Terminal companion to LingCode.app. Run lingcode ask from any shell — routes to the running app, or falls back to DeepSeek / Claude headless when the app is closed.
Privacy policy
What LingCode sends over the network, where API keys live, and why we have zero telemetry.