</>MCP Agents Market
Agent

opensquilla

by opensquilla6.8kPythonUpdated 2026-08-29

OpenSquilla — Token-Efficient AI Agent with same budget, higher intelligence density

OpenSquilla is a token-efficient AI agent built on a microkernel architecture that optimizes cost by routing each conversation turn to the most cost-effective capable model. It provides a unified turn loop accessible through command-line interface, web control console, and multiple chat platforms (Slack, Discord, Telegram, Feishu, Matrix, and more). The agent includes persistent memory, layered sandbox execution, integrated web search, on-device embeddings, and a pluggable provider layer supporting 20+ LLM backends including OpenAI, Anthropic, OpenRouter, Ollama, DeepSeek, Gemini, and Qwen. OpenSquilla's SquillaRouter uses local intelligence to achieve higher capability density within the same budget by dynamically selecting models based on task requirements.

Key Features

Intelligent local model router (SquillaRouter) that selects the cheapest capable model per conversation turn
Unified turn loop shared across CLI, Web UI, and chat channels with consistent tool dispatch and retry behavior
Support for 20+ LLM providers through pluggable architecture (TokenRhythm, OpenRouter, OpenAI, Anthropic, Ollama, DeepSeek, Gemini, Qwen/DashScope)
Multi-channel deployment to Slack, Discord, Telegram, Feishu, DingTalk, QQ, WeCom, and Matrix
Built-in web search integration with DuckDuckGo and Exa providers
Persistent memory system with on-device embeddings for context retention
Cross-platform desktop applications for Windows and macOS with Electron-based control console
Event streaming for subprocess integration with privacy-bounded JSONL progress events

Use Cases

  • 01Cost-optimized conversational AI deployments that automatically route simple queries to cheaper models
  • 02Multi-platform chat agent accessible through Slack, Discord, Telegram, or other team communication tools
  • 03Development teams needing consistent agent behavior across CLI automation and interactive web interfaces
  • 04Organizations requiring on-premise AI agents with local model routing and persistent memory
  • 05Research and experimentation with multi-model ensemble routing strategies
  • 06Automated task execution with event-stream integration for monitoring and logging

Related Agents

View more

opensquilla — FAQ

What is OpenSquilla?+

OpenSquilla is a token-efficient AI agent with a microkernel architecture that routes conversation turns to the most cost-effective model capable of handling each request. It provides CLI, Web UI, and chat channel interfaces through a single shared turn loop with persistent memory and integrated search.

How do I install OpenSquilla?+

The recommended installation uses uv: run 'uv tool install --python 3.12 "opensquilla[recommended] @ <release-wheel-URL>"' or download platform-specific desktop installers (.dmg for macOS, .exe for Windows) from the GitHub releases page. After installation, run 'opensquilla onboard' to configure your LLM provider and settings.

Which LLM providers does OpenSquilla support?+

OpenSquilla supports 20+ providers including TokenRhythm, OpenRouter, OpenAI, Anthropic, Ollama, DeepSeek, Gemini, and Qwen/DashScope through its pluggable provider layer. You configure your preferred provider during the onboarding wizard or via 'opensquilla configure provider'.

Do I need API keys to use OpenSquilla?+

Yes, you need an API key for at least one supported LLM provider. You can use local models via Ollama without cloud API keys, or configure cloud providers like OpenAI, Anthropic, or OpenRouter by setting their API keys during onboarding with '--api-key-env'.

Is OpenSquilla free to use?+

OpenSquilla itself is open source under the Apache 2.0 license and free to use. However, you will incur costs from your chosen LLM provider based on token usage; OpenSquilla's router aims to minimize these costs by selecting cheaper models when appropriate.

What are the system requirements?+

OpenSquilla requires Python 3.12 or later. Desktop installers bundle all dependencies; terminal installs need uv (installed automatically if missing). The SquillaRouter profile requires Visual C++ runtime on Windows and may need libomp via Homebrew on macOS; the agent continues with direct routing if these are unavailable.

How do I install opensquilla?+

Open the source repository on GitHub and follow its README. opensquilla is a agent — MCP Agents Market links you directly to the official repo.

Is opensquilla free?+

opensquilla is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.

Related searches