{"slug": "icecode-self-hosted-ai-agent-platform-multi-agent-swarm-local-rag-26-page-web-ui", "title": "ICECODE — Self-hosted AI Agent Platform: Multi-Agent Swarm, Local RAG, 26-Page Web UI, 34 Platform Gateways", "summary": "ICECODE is a self-hosted AI agent platform that integrates three open-source projects (Hermes, OpenCode, ClawX) into a single system running entirely on a local machine, ensuring no data leaves the user's network. It features a multi-agent swarm with pipeline and parallel orchestration modes, local RAG capabilities, a 26-page web UI, and support for 34 platform gateways. The platform includes a cost optimizer with semantic caching, context compression, and a smart model router, and requires no API keys to operate.", "body_md": "ICECODE is a self-hosted AI agent platform I built by unifying three open-source projects (Hermes, OpenCode, ClawX) into one cohesive system. It runs entirely on your machine — no data leaves your network.\n\n## What it does\n\n-\n**Multi-Agent Swarm**— pipeline and parallel orchestration: agents collaborate, pass context, and vote on answers -** Local RAG**— FAISS + sentence-transformers, 100% offline — index any file, search semantically -** 26-Page Web UI**— dashboard, chat, kanban, goals, swarm, knowledge, MCP, benchmark, and more -** Cost Optimizer**— semantic cache (skip duplicate API calls), context compressor, smart model router -** 174 REST API routes** with auto-docs (FastAPI) -**34 platform gateways**— WhatsApp, Telegram, Discord, Slack, Email, WeChat, Feishu, and 27 more -** 14 LLM providers**— Anthropic, OpenAI, Ollama, Gemini, Mistral, Bedrock, Azure, OpenRouter... -** 90+ agent tools**— file, web, browser, terminal, vision, kanban, MCP, code execution -** Self-learning skills**— agents learn new skills at runtime, stored across sessions -** Reinforcement learning** environment for agent improvement -**WebSocket chat**— bidirectional with cancel support -** Token tracking + cost**— per-session usage across all providers\n\n## Quick start\n\n[1m\n\n╔══════════════════════════════════════════════════════╗\n\n║ ICECODE Super-Agent Network — Installer ║\n\n║ v2.0.0 ║\n\n╚══════════════════════════════════════════════════════╝\n\n[0m\n\n[0;33m→[0m Checking Python version...\n\n[0;32m✓[0m Python 3.12 found\n\n[0;32m✓[0m Virtual environment already exists\n\n[0;33m→[0m Installing Python dependencies...\n\nStarting ICECODE server on [http://localhost:13210](http://localhost:13210)...\n\n[32m23:36:33[0m | [1mINFO [0m | React UI served at /desktop/\n\nINFO: Started server process [423207]\n\nINFO: Waiting for application startup.\n\n[32m23:36:33[0m | [1mINFO [0m | ============================================================\n\n[32m23:36:33[0m | [1mINFO [0m | ICECODE Super-Agent Network v2.0.0 starting...\n\n[32m23:36:33[0m | [1mINFO [0m | Port: 13210\n\n[32m23:36:33[0m | [1mINFO [0m | DB: ~/.icecode/data/icecode.db\n\n[32m23:36:33[0m | [1mINFO [0m | Home: /home/claudiu/.icecode\n\n[32m23:36:33[0m | [1mINFO [0m | ============================================================\n\n[32m23:36:33[0m | [1mINFO [0m | [✓] Database initialized at ~/.icecode/data/icecode.db\n\n[32m23:36:33[0m | [1mINFO [0m | [✓] Self-learning system ready\n\n[32m23:36:33[0m | [1mINFO [0m | [✓] Cron scheduler ready\n\n[32m23:36:33[0m | [1mINFO [0m | [✓] Goals system (Ralph Loop) ready\n\n[32m23:36:33[0m | [1mINFO [0m | [✓] Knowledge auto-index task started\n\n[32m23:36:33[0m | [1mINFO [0m | All ICECODE systems online.\n\nINFO: Application startup complete.\n\nERROR: [Errno 98] error while attempting to bind on address ('0.0.0.0', 13210): address already in use\n\nINFO: Waiting for application shutdown.\n\n[32m23:36:33[0m | [1mINFO [0m | ICECODE shutting down...\n\nINFO: Application shutdown complete.\n\n## Architecture\n\n## Cost Optimizer (new in v2)\n\nThe cost optimizer has three components that work together automatically:**Semantic Cache**— uses sentence-transformer embeddings + cosine similarity (≥0.92 threshold). If you ask a semantically similar question to a previous one, it returns the cached answer instantly — zero API tokens consumed.**Context Compressor**— when conversation history exceeds 3000 tokens, older messages are summarized instead of sent verbatim. Keeps the last 6 messages intact.**Smart Model Router**— analyzes prompt complexity (score 1-10) and routes to the cheapest capable model. Simple questions go to cheap models, complex reasoning goes to powerful ones.\n\n## Multi-Agent Swarm\n\nTwo orchestration modes:**Pipeline**: agents run sequentially, each building on the previous output.** Parallel**: all agents receive the same input simultaneously, results are merged.\n\nBuilt-in templates: Research & Write, Code Review, Brainstorm, Security Audit.\n\n## Local RAG\n\nNo API keys needed. No data leaves your machine.\n\n- Supports , , , , , , ,\n- Chunk strategy: sliding window (512 tokens, 50 overlap)\n- Embeddings: (90MB, runs locally)\n- Vector store: FAISS IndexFlatL2\n- Persistence:\n\nIndex a directory:\n\n{\"detail\":[{\"type\":\"json_invalid\",\"loc\":[\"body\",0],\"msg\":\"JSON decode error\",\"input\":{},\"ctx\":{\"error\":\"Expecting value\"}}]}\n\n## Tech stack**Backend**— Python 3.12 · FastAPI · Uvicorn · SQLite · Pydantic v2 · FAISS · sentence-transformers · Loguru** Frontend**— Single HTML file, no build step · Pure JS ES2022 · CSS variables · Server-Sent Events + WebSocket** TypeScript**— pnpm workspaces · turbo · CLI with Ink TUI · 14 LLM provider protocols · MCP client** Desktop**— Electron 33** Infrastructure** — Docker · GitHub Actions CI · pytest (108 tests) · ruff\n\n## Links\n\n- GitHub:\n[https://github.com/iceslim409/icecode](https://github.com/iceslim409/icecode) - Demo GIF in README shows all major pages\n- MIT-style non-commercial license (ICECODE-NC-1.0)\n\nIf you try it, I'd love to hear what you think. Issues and PRs welcome.", "url": "https://wpnews.pro/news/icecode-self-hosted-ai-agent-platform-multi-agent-swarm-local-rag-26-page-web-ui", "canonical_source": "https://dev.to/soros_02159c95a2582796088/icecode-self-hosted-ai-agent-platform-multi-agent-swarm-local-rag-26-page-web-ui-34-platform-ig3", "published_at": "2026-05-21 05:36:50+00:00", "updated_at": "2026-05-21 06:21:49.175358+00:00", "lang": "en", "topics": ["artificial-intelligence", "open-source", "developer-tools", "products"], "entities": ["ICECODE", "Hermes", "OpenCode", "ClawX"], "alternates": {"html": "https://wpnews.pro/news/icecode-self-hosted-ai-agent-platform-multi-agent-swarm-local-rag-26-page-web-ui", "markdown": "https://wpnews.pro/news/icecode-self-hosted-ai-agent-platform-multi-agent-swarm-local-rag-26-page-web-ui.md", "text": "https://wpnews.pro/news/icecode-self-hosted-ai-agent-platform-multi-agent-swarm-local-rag-26-page-web-ui.txt", "jsonld": "https://wpnews.pro/news/icecode-self-hosted-ai-agent-platform-multi-agent-swarm-local-rag-26-page-web-ui.jsonld"}}