{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreihrsgf7vzyrr7vrulq6div4n6kzotnpx4hjuq4yntrqwm4fkzuv6a",
    "uri": "at://did:plc:25rdn5elo5izoxrmtis34zuk/app.bsky.feed.post/3molyu5flqm42"
  },
  "coverImage": {
    "$type": "blob",
    "ref": {
      "$link": "bafkreien3zsenrtxhkrnzp5h4swyr54mqgoonufio4p2eiwkaihq7wtawy"
    },
    "mimeType": "image/webp",
    "size": 74090
  },
  "path": "/omni_loopresearchlabs/clioloop-an-open-source-ai-agent-with-agentic-fusion-3lk9",
  "publishedAt": "2026-06-18T22:56:12.000Z",
  "site": "https://dev.to",
  "tags": [
    "ai",
    "opensource",
    "python",
    "agents",
    "https://github.com/Clioloop/Clioloop-agent",
    "https://portal.clioloop.com",
    "https://portal.clioloop.com/docs"
  ],
  "textContent": "##  The problem with single-model AI assistants\n\nMost AI assistants give you one model's answer. If it's wrong, you catch it or you don't. If you use a cheap model, quality drops. If you use a frontier model, you pay frontier prices for everything — even a simple file rename.\n\nWe wanted something better. So we built Clioloop.\n\n##  What is Agentic Fusion?\n\nAgentic Fusion puts a whole team of models on one task. When you run `/fusion`:\n\n  1. **Planners** (up to 5): Read-only models that research and propose routes in parallel. They figure out the best approach but can't touch your files or run commands.\n\n  2. **Main model** : Your chosen model does the actual work — full tool access, fully visible. You watch every step. Not a black box.\n\n  3. **Reviewers** (up to 5): Read-only models that critique the draft. They can see images the main model generated. They check for errors, suggest fixes, flag issues.\n\n  4. **Verdict loop** : The draft is revised until reviewers approve. The answer you get has already passed independent review.\n\n  5. **Fusion** : Everything combines into one reviewed, approved answer.\n\n\n\n\n###  The key insight\n\nThe quality comes from **synthesis** — not from running the same job 5 times. Cheap open models combine into something that rivals a frontier model, at a fraction of the cost.\n\n###  Safety by construction\n\nPlanners and reviewers are read-only at the schema level. They can research and critique, but they can never touch your files or execute commands. Only your main model has tool access, and you watch it work live.\n\n##  Beyond fusion\n\nClioloop is also:\n\n  * **Self-improving** : Keeps `MEMORY.md` and `USER.md`, updated automatically as it learns your preferences and projects\n  * **Autonomous** : Set a standing goal with `/goal` and it loops — planning, running tools, checking progress — until done\n  * **Everywhere** : Terminal, desktop app, web dashboard, Telegram, Slack, Discord, WhatsApp — same session, same memory\n  * **Multi-agent Kanban** : Break big work into a board of tasks; worker agents pick them up and report back\n  * **Tools** : File editing, shell, web search & extract, cloud browser, image generation (ComfyUI/Flux), video generation, TTS, any MCP server\n  * **Scheduled jobs** : Run it on cron for automated workflows\n  * **Open-source** : Self-host everything. Own your data. No vendor lock-in.\n\n\n\n##  The Omni Loop Portal\n\nOne OAuth login gives you access to 300+ models. No API keys to manage. An OpenAI-compatible proxy means you can point any existing tool at it.\n\nThe tool gateway handles:\n\n  * Web search & extract (Firecrawl)\n  * Browser automation (Browser Use)\n  * Image generation (FAL/Flux)\n  * Video generation (Vidu)\n  * Text-to-speech (OpenAI TTS)\n\n\n\nAll metered, all behind one account.\n\n##  Install\n\n\n    # macOS / Linux\n    curl -fsSL https://raw.githubusercontent.com/Clioloop/Clioloop-agent/main/scripts/install.sh | bash\n\n    # Windows (PowerShell)\n    iex (irm https://raw.githubusercontent.com/Clioloop/Clioloop-agent/main/scripts/install.ps1)\n\n\nThen:\n\n\n\n    clio setup    # pick Omni Loop Portal, approve in browser\n    clio          # start looping\n\n\n##  Architecture overview\n\nThe repo is a monorepo:\n\n  * **Python core** (`clio` CLI + agent loop in `clio_cli/` and `agent/`)\n  * **TypeScript surfaces** (Electron desktop, React-Ink TUI, React web dashboard)\n  * **Omni Loop Portal** (Next.js 15 + SQLite + Stripe)\n\n\n\nEvery provider is a `ProviderProfile` with lazy discovery. The `managed` provider _is_ the Omni Loop Portal (OAuth device-code, no API key). Model catalog + switching is cached locally.\n\nThe fusion engine lives in `agent/fusion_engine.py`. Slash commands have a single source of truth in `clio_cli/commands.py`.\n\n##  Links\n\n  * GitHub: https://github.com/Clioloop/Clioloop-agent\n  * Portal: https://portal.clioloop.com\n  * Docs: https://portal.clioloop.com/docs\n\n\n\nWe're building this in the open. PRs, issues, feedback — all welcome.\n\nWhat would you want from an AI agent? Let us know in the comments.",
  "title": "Clioloop: An Open-Source AI Agent with Agentic Fusion"
}