Item detail

opensquilla/opensquilla

OpenSquilla is an Apache-2.0 'token-efficient' AI agent microkernel with a local model router that picks the cheapest model that can handle each turn, a layered sandbox, persistent memory, built-in web search, and on-device embeddings - all behind a single shared turn loop that the Web UI, CLI, and chat channels all run through. v0.3.1 speaks OpenRouter, OpenAI, Anthropic, Ollama, DeepSeek, Gemini

Score8.0
Popularity90.0
Risknone
TierGold
Score breakdown
Usefulness8.0
Novelty8.0
Momentum9.0
Maturity8.2
Open-source/build8.4
Evidence7.2
Workflow potential9.1
Setup ease8.8

Popularity is tracked separately. Support, ads, sponsorships, and tips never affect these signals.

Why it matters

Useful for developers paying for premium model tokens on tasks a cheaper model already does well. Install via the Windows portable / macOS DMG / pip / source paths, point OpenSquilla at your existing OpenAI / Anthropic / Ollama / DeepSeek keys, and watch the router send 80% of routine turns to a cheap model while reserving the expensive one for the parts that actually need it.

Who should use it

developers paying for premium model tokens on tasks a cheaper model already does well and want a router to pick automaticallyteams who want a single agent runtime that speaks OpenAI / Anthropic / Ollama / DeepSeek / Gemini / Qwen without lock-inprivacy-sensitive workflows that need on-device embeddings and a local model fallback for routine turnsdevelopers who want Windows portable / macOS DMG / pip / source install paths instead of a Linux-only projectanyone building a chat-channel agent (Telegram, Slack, Discord) who needs the same turn loop on Web UI, CLI, and chat

Who should skip it

Skip if the source link, docs, or setup requirements do not match your workflow.

Risk explanation

No inherent user-impacting risk is flagged from the captured evidence.

Evidence links

Closest alternatives / related signals

agentmicrokerneltoken-efficientmodel-routersandboxmemoryon-device-embeddingsopenrouter