From liteharness
Canonical architecture reference for LiteSuite mega-app and ecosystem. Quick-lookup index into 15 architecture docs covering panels, services, MCP tools, voice, harness, arena, and more. Triggers on 'arch', 'architecture', 'show architecture', 'arch docs'.
How this skill is triggered — by the user, by Claude, or both
Slash command
/liteharness:ls-archThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Quick-lookup index into the **canonical** architecture docs for the consolidated LiteSuite mega-app.
Quick-lookup index into the canonical architecture docs for the consolidated LiteSuite mega-app.
Note: This skill reads architecture docs from your project's
docs/architecture/directory. If you are working on LiteSuite, these docs live in your LiteSuite project root. For other projects, use/arch-gento generate architecture docs first.
| # | File | Subject |
|---|---|---|
| 00 | 00-Ecosystem-Overview.md | 3-app ecosystem, port map, consolidation history, directory layout |
| 01 | 01-LiteSuite-Architecture.md | Monorepo, Electron main, IPC (44 handlers), services, preloads |
| 02 | 02-Panel-System.md | 21+ pane types, sidebar morphing, canvas/zen, voice nav (48 cmds) |
| 03 | 03-Backend-Server.md | Bun/Effect WebSocket, protocol, SQLite, orchestration |
| 04 | 04-Voice-Pipeline.md | STT/TTS, wake detection, voice commands, companion, emotion |
| 05 | 05-LiteHarness.md | 5-tier orchestration, git-as-memory, harness MCP tools |
| 06 | 06-MCP-Tools.md | 20 MCP tools, Agent Bridge (:7423), just-bash/lite__shell |
| 07 | 07-LiteBench-Arena.md | Gauntlet, arena battles, ELO, eval harness (10 benchmarks) |
| 08 | 08-LiteMemory-LCM.md | LCM DAG, Memory Graph 3D, vault integration |
| 09 | 09-LiteAgent.md | Python CLI, identity, scheduler, evolution engine |
| 10 | 10-LiteImage.md | Standalone GPU: sd.cpp, face swap, avatar cam, video gen |
| 11 | 11-LiteDock.md | Standalone Rust Stage Manager for Windows 11 |
| 12 | 12-litesuite-dev.md | SaaS: Next.js 15, Cloudflare, Stripe, /harness explorer |
| 13 | 13-AgentsOverflow.md | Worldwide agent network, daily builds, voting |
| 14 | 14-Self-Improvement.md | Self-improvement loop, LiteCLI compilation, /train |
docs/architecture/INDEX.md in the project root (search up from cwd if needed)/arch-gen to generate docs firstINDEX.md for the full table of contents + section lookup tableLiteSuite-specific ports (only relevant when LiteSuite is running locally):
| Port | Service |
|---|---|
| 3773 | Backend server (HTTP + WebSocket) |
| 7423 | Agent Bridge (REST, bearer token) — requires LiteSuite running |
| 7426 | LiteImage REST API — requires LiteSuite running |
| 7438 | Voice API server — requires LiteSuite running |
| 5123 | TTS Server (Qwen3, on-demand) — requires LiteSuite running |
| 8080 | Whisper STT (on-demand) — requires LiteSuite running |
WORKSPACE: terminal, unified-editor, browser, files, git AI: frontier-chat, claude, codex, liteagent-chat, claude-teams TOOLS: benchmark, youtube, voice, memory-graph, modelHub SYSTEM: dashboard, agent, settings, design-system, screens, style-test
Provides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub ahostbr/liteharness-plugin --plugin liteharness