freshcrate
Skin:/
Home > MCP Servers > osaurus

osaurus

Own your AI. The native macOS harness for AI agents -- any model, persistent memory, autonomous execution, cryptographic identity. Built in Swift. Fully offline. Open source.

Why this rank:Strong adoptionRecent releaseHealthy release cadence

Description

Own your AI. The native macOS harness for AI agents -- any model, persistent memory, autonomous execution, cryptographic identity. Built in Swift. Fully offline. Open source.

README

Screenshot 2026-03-19 at 3 42 04โ€ฏPM

Osaurus

Own your AI.
Agents, memory, tools, and identity that live on your Mac. Built purely in Swift. Fully offline. Open source.

Release Downloads License Stars Platform OpenAI API Anthropic API Ollama API MCP Server Foundation Models Hugging Face PRs Welcome

Download for Mac ยท Docs ยท Models ยท Discord ยท Twitter ยท Plugin Registry


Inference is all you need. Everything else can be owned by you.

Models are getting cheaper and more interchangeable by the day. What's irreplaceable is the layer around them -- your context, your memory, your tools, your identity. Others keep that layer on their servers. Osaurus keeps it on your machine.

Osaurus is the AI harness for macOS. It sits between you and any model -- local or cloud -- and provides the continuity that makes AI personal: agents that remember, execute autonomously, run real code, and stay reachable from anywhere. The models are interchangeable. The harness is what compounds.

Works fully offline with local models. Connect to any cloud provider when you want more power. Nothing leaves your Mac unless you choose.

Native Swift on Apple Silicon. No Electron. No compromises. MIT licensed.

Install

brew install --cask osaurus

Or download the latest .dmg from Releases. After installing, launch from Spotlight (โŒ˜ Space โ†’ "Osaurus") or the CLI:

osaurus ui       # Open the chat UI
osaurus serve    # Start the server
osaurus status   # Check status

Requires macOS 15.5+ and Apple Silicon.

Agents

Agents are the core of Osaurus. Each one gets its own prompts, memory, and visual theme -- a research assistant, a coding partner, a file organizer, whatever you need. Tools and skills are automatically selected via RAG search based on the task at hand -- no manual configuration needed. Everything else in the harness exists to make agents smarter, faster, and more capable over time.

Agent Loop

Every chat is an agent loop. Pick a working folder and the agent gets file, search, and git tools. Toggle the sandbox and it gets shell access in an isolated Linux VM. The model writes a markdown todo list, executes against it, and closes out with a verified summary -- all in the same chat window. See the Agent Loop Guide.

Sandbox

Agents execute code in an isolated Linux VM powered by Apple's Containerization framework. Full dev environment -- shell, Python, Node.js, compilers, package managers -- with zero risk to your Mac.

Each agent gets its own Linux user and home directory. The VM connects back to Osaurus (inference, memory, secrets) via a vsock bridge -- sandboxed but not disconnected. Extend with simple JSON plugin recipes, no Xcode or code signing required.

โ”Œโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”       โ”Œโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”
โ”‚    Osaurus     โ”‚       โ”‚   Linux VM (Alpine)        โ”‚
โ”‚                โ”‚       โ”‚                            โ”‚
โ”‚  Sandbox Mgr โ”€โ”€โ”ผโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”คโ†’ /workspace  (VirtioFS)    โ”‚
โ”‚  Host API   โ†โ”€โ”€โ”ผโ”€vsockโ”€โ”คโ†’ osaurus-host bridge       โ”‚
โ”‚                โ”‚       โ”‚                            โ”‚
โ”‚                โ”‚       โ”‚  agent-alice  (Linux user) โ”‚
โ”‚                โ”‚       โ”‚  agent-bob    (Linux user) โ”‚
โ””โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”˜       โ””โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”˜

Requires macOS 26+ (Tahoe). See the Sandbox Guide for configuration, built-in tools, and plugin authoring.

Memory

Three layers -- identity, pinned facts, and per-session episodes -- plus a transcript fallback. Agents distill conversations once at session end (not on every turn), score what matters by salience, and surface at most one compact slice per request based on what you're actually asking. A background consolidator decays, merges, and evicts so memory stays sharp instead of bloating. Most turns inject ~800 tokens or less; many inject zero. See the Memory Guide.

Identity

Every participant -- human, agent, device -- gets a secp256k1 cryptographic address. Authority flows from your master key (iCloud Keychain) down to each agent in a verifiable chain of trust. Create portable access keys (osk-v1), scope per-agent, revoke anytime. See Identity docs.

Relay

Expose agents to the internet via secure WebSocket tunnels through agent.osaurus.ai. Unique URL per agent based on its crypto address. No port forwarding, no ngrok, no configuration.

Models

The harness is model-agnostic. Swap freely -- your agents, memory, and tools stay intact.

Local

Run Gemma 4, Qwen3.6, GPT-OSS, Llama, and more on Apple Silicon with optimized MLX inference. Osaurus maintains its own optimized model library on Hugging Face with curated quantizations for the best quality-to-size ratio on Apple Silicon. Models stored at ~/MLXModels (override with OSU_MODELS_DIR). Fully private, fully offline.

Liquid Foundation Models

Osaurus supports Liquid AI's LFM family -- on-device models built on a non-transformer architecture optimized for edge deployment. Fast decode, low memory footprint, and strong tool calling out of the box.

Apple Foundation Models

On macOS 26+, use Apple's on-device model as a first-class provider. Pass model: "foundation" in API requests. Tool calling maps through Apple's native interface automatically. Zero inference cost, fully private.

Cloud

Connect to OpenAI, Anthropic, Gemini, xAI/Grok, Venice AI, OpenRouter, Ollama, or LM Studio. Venice provides uncensored, privacy-focused inference with no data retention. Context and memory persist across all providers.

MCP

Osaurus is a full MCP (Model Context Protocol) server. Give Cursor, Claude Desktop, or any MCP client access to your tools:

{
  "mcpServers": {
    "osaurus": {
      "command": "osaurus",
      "args": ["mcp"]
    }
  }
}

Also an MCP client -- aggregate tools from remote MCP servers into Osaurus. See the Remote MCP Providers Guide for details.

Tools & Plugins

osaurus tools install osaurus.browser    # Install from registry
osaurus tools list                       # List installed
osaurus tools create MyPlugin --swift    # Create a plugin
osaurus tools dev com.acme.my-plugin     # Dev with hot reload

20+ native plugins: Mail, Calendar, Vision, macOS Use, XLSX, PPTX, Browser, Music, Git, Filesystem, Search, Fetch, and more. Plugins support v1 (tools only) and v2 (full host API) ABIs -- register HTTP routes, serve web apps, persist data in SQLite, dispatch agent tasks, and call inference through any model. See the Plugin Authoring Guide.

More

Skills & Methods -- Skills import reusable AI capabilities from GitHub repos or files, compatible with Agent Skills. Methods are learned workflows that agents save and reuse over time. Both are automatically selected via RAG search -- no manual configuration needed. See Skills Guide.

Automation -- Schedules run recurring tasks in the background. Watchers monitor folders and trigger agents on file changes.

Voice -- On-device transcription via FluidAudio on Apple's Neural Engine. Voice input in chat, VAD mode with wake-word activation, and a global hotkey to transcribe into any app. No audio leaves your Mac. See Voice Input Guide.

Developer Tools -- Server explorer, MCP tool inspector, inference monitoring, plugin debugging. See Developer Tools Guide. For the inference scheduler, model leases, continuous-batching engine, and feature flags that tune them, see Inference Runtime.

Compatible APIs

Drop-in endpoints for existing tools:

API Endpoint
OpenAI http://127.0.0.1:1337/v1/chat/completions
Anthropic http://127.0.0.1:1337/anthropic/v1/messages
Ollama http://127.0.0.1:1337/api/chat

All prefixes supported (/v1, /api, /v1/api). Full function calling with streaming tool call deltas. See OpenAI API Guide for tool calling, streaming, and SDK examples. Building a macOS app that connects to Osaurus? See the Shared Configuration Guide.

CLI

osaurus serve --port 1337              # Start on localhost
osaurus serve --port 1337 --expose     # Expose on LAN
osaurus ui                             # Open the chat UI
osaurus status                         # Check status
osaurus stop                           # Stop the server

Homebrew auto-links the CLI, or symlink manually:

ln -sf "/Applications/Osaurus.app/Contents/MacOS/osaurus" "$(brew --prefix)/bin/osaurus"

Architecture

โ”Œโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”
โ”‚                   The Harness                       โ”‚
โ”œโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ฌโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ฌโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ฌโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ค
โ”‚ Agents   โ”‚ Memory   โ”‚ Agent Loop โ”‚ Automation       โ”‚
โ”œโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ดโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ดโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ดโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ค
โ”‚              MCP Server + Client                    โ”‚
โ”œโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ฌโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ฌโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ฌโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ค
โ”‚ MLX      โ”‚ OpenAI   โ”‚ Anthropic โ”‚ Ollama / Others   โ”‚
โ”‚ Runtime  โ”‚ API      โ”‚ API       โ”‚                   โ”‚
โ”œโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ดโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ดโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ดโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ค
โ”‚      Plugin System (v1 / v2 ABI) ยท Native Plugins   โ”‚
โ”œโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ฌโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ฌโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ฌโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ค
โ”‚ Identity โ”‚ Relay    โ”‚ Tools     โ”‚ Skills ยท Methods  โ”‚
โ”œโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ดโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ดโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ดโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”ค
โ”‚  Sandbox VM (Alpine ยท Apple Containerization)       โ”‚
โ”‚  vsock bridge ยท VirtioFS ยท per-agent isolation      โ”‚
โ””โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”˜

Most features are accessible through the Management window (โŒ˜ โ‡ง M).

Build from Source

git clone https://github.com/osaurus-ai/osaurus.git
cd osaurus
open osaurus.xcworkspace

Build and run the osaurus target. Requires Xcode 16+ and macOS 15.5+.

Git Hooks (lefthook)

Install lefthook to set up the hooks that verify quality of the code:

brew install lefthook
lefthook install

This installs a pre-push hook that runs swift-format over the Packages/ directory before each push.

Project Structure

osaurus/
โ”œโ”€โ”€ App/                          # macOS app target (SwiftUI entry point, assets, entitlements)
โ”œโ”€โ”€ Packages/
โ”‚   โ”œโ”€โ”€ OsaurusCore/              # Core library โ€” all app logic
โ”‚   โ”‚   โ”œโ”€โ”€ Models/               # Data types, DTOs, configuration stores
โ”‚   โ”‚   โ”œโ”€โ”€ Services/             # Business logic (actors and stateless types)
โ”‚   โ”‚   โ”œโ”€โ”€ Managers/             # UI-facing state holders (@MainActor, observable)
โ”‚   โ”‚   โ”œโ”€โ”€ Views/                # SwiftUI views, organized by feature
โ”‚   โ”‚   โ”œโ”€โ”€ Networking/           # HTTP server, routing, relay
โ”‚   โ”‚   โ”œโ”€โ”€ Storage/              # SQLite databases
โ”‚   โ”‚   โ”œโ”€โ”€ Identity/             # Cryptographic identity and access keys
โ”‚   โ”‚   โ”œโ”€โ”€ Tools/                # MCP tools, plugin ABI, tool registry
โ”‚   โ”‚   โ”œโ”€โ”€ Folder/               # Working-folder context, file ops, batch tool
โ”‚   โ”‚   โ”œโ”€โ”€ Utils/                # Cross-cutting utilities
โ”‚   โ”‚   โ””โ”€โ”€ Tests/                # Unit and integration tests
โ”‚   โ”œโ”€โ”€ OsaurusCLI/               # CLI (osaurus command)
โ”‚   โ””โ”€โ”€ OsaurusRepository/        # Plugin registry and installation
โ”œโ”€โ”€ docs/                         # Feature guides and documentation
โ”œโ”€โ”€ scripts/                      # Build, release, and benchmark scripts
โ”œโ”€โ”€ sandbox/                      # Sandbox VM Dockerfile
โ””โ”€โ”€ assets/                       # DMG packaging assets

See CONTRIBUTING.md for the architecture guide and layer definitions.

Contributing

Osaurus is actively developed and we welcome contributions: bug fixes, new plugins, documentation, UI/UX improvements, and testing.

Check out Good First Issues, read the Contributing Guide, or join Discord. See docs/FEATURES.md for the full feature inventory.

Community

License

MIT


Osaurus, Inc. ยท osaurus.ai

Release History

VersionChangesUrgencyDate
0.25.26## What's Changed - Raise release build timeout to 180 minutes (#3181) by @tpae - Add opt-in local access key for sensitive routes, settable by bundle import (#3178) by @tpae - Harness runtimes in the model picker, config documents, and bundles (#3180) by @tpae - Add pluggable harness runtimes: Claude Code and Pi (#3175) by @tpae - Add `osaurus import` for Osaurus bundles (#3174) by @tpae - Add agent templates (file export/import) (#3172) by @tpae ## ๐Ÿ› Bug Fixes - Fix: Prevent main thread haHigh10/24/2026
0.25.25## What's Changed - feat: integrate Flint runtime, owned weights and SSD reuse fixes (#3161) by @jjang-ai - Remove retired cold-weight tier setting and status APIs (#3156) by @jjang-ai - Align adaptive drafter loading and remove legacy fixed widths (#3150) by @jjang-ai - Pin hybrid cache checkpoint RAM improvements (#3147) by @jjang-ai ## ๐Ÿ› Bug Fixes - Pin lower batch checkpoint memory and shared SSD writer reservations (#3168) by @jjang-ai - Resolve declared reasoning tiers across bundle meHigh10/22/2026
0.25.24## What's Changed - FLUX 3 Image Klein: generate, edit, references, delegated jobs (#3145) by @jjang-ai - Pin vmlx-swift #623: Ornith 122B Adaptive MTP by default, colour parity, router tail (#3142) by @jjang-ai - Ornith 122B: architecture prefill step, pin vmlx-swift#621 cache reuse (#3140) by @jjang-ai - make phone runs own their mac chat, catch spoken pii and let a phone call set the privacy filter (#3135) by @RaajeevChandran - Allosaurus 0.2 on 48 GB Macs: pin vmlx-swift adaptive-MTP lookahHigh10/20/2026
0.25.23## What's Changed - Integrate native workflow media controls, video and memory admission (#3104) by @jjang-ai - Nemotron 3.5 Super-VL and Bolt runtime: vmlx-swift pin and declared reasoning modes (#3119) by @jjang-ai - ask for a short spoken reply on a phone's voice-call turns, as a note after the conversation that is never saved (#3106) by @RaajeevChandran - Consume native image batches across API tools and galleries (#3098) by @jjang-ai - Expose SeedVR2 upscaling and cooperative job cancellatHigh10/17/2026
0.25.22## What's Changed ## ๐Ÿ› Bug Fixes - fix(runtime): pin Flash-Next MTP speed restoration (vmlx-swift #593) (#3087) by @jjang-ai - Count all production compute weights in model and subagent admission (#3085) by @jjang-ai - Propagate screenshot scrubbing failures before cloud attachment (#3084) by @jjang-ai - Rank pinned memory text matches by relevance before salience (#3083) by @jjang-ai - Integrate native Liquid tools and corrected LFM affine/JANGH runtime (#3082) by @jjang-ai - Chat table: coaHigh10/14/2026
0.25.21## What's Changed - MiniCPM-V 4.7 + DSpark: pin vmlx-swift, DSpark load warmup, solo lease for drafter requests (#3075) by @jjang-ai - Pin vmlx-swift #572: background cache writes (shorter Finishing up) and bundled Uno for K2 (#3063) by @jjang-ai - stop showing injected memory in user bubbles for chats continued from the phone (#3070) by @RaajeevChandran - fix /api/show 404 for external models listed by repo name (#3069) by @htjworld - Support Qwen-Image-2.1-Turbo image bundles (#3059) by @jjanHigh10/13/2026
0.25.20## What's Changed - Add calculate built-in tool for exact math (#3039) by @jjang-ai - Say why images are unavailable instead of dropping a pasted image (#3021) (#3040) by @jjang-ai - harden the phone's secure channel and pairing, and expand phone image generation (#3044) by @RaajeevChandran - Consume merged tool-stream and K2 runtime fixes (#3046) by @jjang-ai - Clarify judge polarity and ScreenContext rubric while preserving verdict gating (#3033) by @jjang-ai - Compaction turns reasoning off High10/8/2026
0.25.19## What's Changed - Unify chat composer popups with the model picker style and quiet the lone chat tab (#3017) by @tpae - let the paired phone manage its agents and chats and give it a secure channel on a mac with no custom agents (#2980) by @RaajeevChandran - Pin qualified GLM required-tool cache restoration (#3002) by @jjang-ai ## ๐Ÿš€ Features - steer channel messages into the running turn instead of waiting for the next one (#3010) by @RaajeevChandran ## ๐Ÿ› Bug Fixes - Fix non-MLX Sentry High10/5/2026
0.25.18## What's Changed - Support Qwen-Image-2.1 generation and editing (#2997) by @jjang-ai - Restyle chat tab strip as a Safari-style track (#2995) by @tpae - Pin Qwen4Exp JANGH runtime and media cache fixes (#2991) by @jjang-ai - MCP robustness, elicitation, and regulated-industry connectors (#2990) by @tpae - Use exact Naive rotary application on qualified M5 Max (#2972) by @jjang-ai ## ๐Ÿ› Bug Fixes - Pin GLM sparse-indexer device range optimization (#3001) by @jjang-ai - Pin canonical Qwen staHigh10/4/2026
0.25.17## What's Changed - PDF reading-order extraction for flattened forms (file\_read pages, page markers) (#2983) by @tpae - Router announcements feed and control-plane polling cuts (#2982) by @tpae - International privacy filter presets with region picker (#2979) by @tpae - Insights: tamper-evident activity log, Tier 1 coverage gaps, and IA redesign (#2964) by @tpae - Consume Naive attention-mask GPU ranges (#2967) by @jjang-ai - Seed Osaurus Cloud starter favorites and unify picker Model options High10/3/2026
0.25.16## What's Changed - Pin MLX 0.32.3 for the crash-fix release (#2956) by @jjang-ai - Redesign chat model picker, Cloud browser and Credits menu (#2947) by @losingmyego - Settings UX/IA cleanup: grouped-form kit, renamed tabs, Tools \& MCP Services rework (#2950) by @tpae ## ๐Ÿ› Bug Fixes - Pin runtime crash containment and allocator synchronization fixes (#2953) by @jjang-ai - Reject invalid process strings before Foundation launch (#2954) by @jjang-ai - only show the model switch advisory whenHigh10/1/2026
0.25.15## What's Changed - Orchestrator audit remediation: truthful addendum, apply-first config, pool guard, default new-chat agent, compact spawn envelope (#2936) by @tpae - keep the phone's runs going when its connection drops and let it rejoin or stop them by run id (#2935) by @RaajeevChandran - Pin the synchronized ErrorBox runtime (#2934) by @jjang-ai - bring phone chats to the front on return to the mac and send the orchestrator's quick actions to the phone (#2930) by @RaajeevChandran - send thHigh9/29/2026
0.25.14## What's Changed - Clarify background model admission for description backfill (#2901) by @jjang-ai - added iOS app integrations (#2875) by @RaajeevChandran ## ๐Ÿ› Bug Fixes - Fix covariant Self default argument in ProductHuntLaunchCampaign init (#2906) by @tpae - Fix timing export choices for restored chats (#2902) by @jjang-ai - Preserve generation metrics across chat history reloads (#2900) by @jjang-ai - fix right-click copy dropping the chat drag selection (#2899) by @RaajeevChandran - SHigh9/27/2026
0.25.13## What's Changed - Make agent description optional again with background backfill (#2892) by @tpae - Retire Raptor v0.5 and shorten the What's New notes (#2882) by @tpae - Make cancellation and runtime-settings tests deterministic (#2885) by @jjang-ai - Reuse canonical prefill checkpoints and preserve current model discovery (#2884) by @jjang-ai - Use qualified Raptor q6 decode fusion (#2883) by @jjang-ai - Reuse completed prefill chunks in batched cache finalization (#2881) by @jjang-ai - FusHigh9/26/2026
0.25.12## What's Changed ## ๐Ÿ› Bug Fixes - Fall back from an unavailable or hung Foundation core model (#2861) by @tpae - Decide Automation permission probes from the error code, not the localized text (#2858) (#2859) by @tpae ## ๐Ÿงฐ Maintenance - Move the Raptor Product Hunt launch to Monday (#2860) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.25.11...0.25.12 High9/23/2026
0.25.9## What's Changed - Recognize GPT-6 Astra and other new remote-provider models (#2826) by @tpae - Hide dispatch envelopes behind a compact source badge row (#2824) by @tpae ## ๐Ÿ› Bug Fixes - Fix app hang in SelectableNSTextView.hitTest (#2827) by @[sentry[bot]](https://github.com/apps/sentry) - Show n8n pending approval card with preset allowlists (#2829) by @tpae - fix app hang from localization lookups in liveStatusLine (#2812) by @[sentry[bot]](https://github.com/apps/sentry) - Preserve reHigh9/19/2026
0.25.8## What's Changed - Pin Bonsai FP16 attention K/V with FP32 recurrent state (#2820) by @jjang-ai - Consume merged Bonsai loading and bounded media prefill (#2818) by @jjang-ai ## ๐Ÿ› Bug Fixes - Sandbox: attribute provisioning failures and stop the quit-time VM re-boot behind vmnet\_in\_use (#2819) by @tpae - Preserve persisted document text in chat and warmup (#2816) by @jjang-ai - fixed main thread hangs in the composer, markdown parsing and sidebar rows (#2817) by @RaajeevChandran - Share pHigh9/18/2026
0.25.7## What's Changed - Embed the Workspaces explainer in the tab's empty state; drop the intro dialog (#2811) by @tpae - Support Bonsai2 runtime and preserve native Qwen tool templates (#2808) by @jjang-ai ## ๐Ÿš€ Features - added number\_of\_agents property to agent\_created telemetry event (#2807) by @RaajeevChandran ## ๐Ÿ› Bug Fixes - Preserve typed MCP image results in Chat and agent history (#2805) by @jjang-ai - fixed tab strip jumping and overlapping the sidebar on window resize (#2802) byHigh9/18/2026
0.25.5## What's Changed - Orchestrator polish and simplification (#2787) by @tpae - Default onboarding to Raptor 0.6 (#2782) by @tpae ## ๐Ÿ› Bug Fixes - Make model repair verify files and report download progress (#2788) by @jjang-ai - Preserve local usage accounting and complete image chats cleanly (#2786) by @jjang-ai - Keep MTP opt-in and enforce runtime settings across Chat and API (#2785) by @jjang-ai - Fix resident child RAM admission and external model residency tracking (#2784) by @jjang-ai High9/16/2026
0.25.4## What's Changed - Remote prompt-cache efficiency: session cache keys, cache-aware Router telemetry, BYOK cache parsing (#2779) by @tpae - Guide n8n setup so pairing codes only issue when they can work (#2775) by @tpae - QoL improvements (#2776) by @RaajeevChandran - Audit installed vision bundles and repair discovery and admission parity (#2772) by @jjang-ai - Multi-device identity: unique agent addresses, owner redeem, and iOS Keychain group (#2769) by @tpae ## ๐Ÿš€ Features - add recent folHigh9/15/2026
0.25.3## What's Changed - Add n8n pairing code for remote Secure Channel setup (#2762) by @tpae - qol improvements for providers and the chat sidebar (#2755) by @RaajeevChandran ## ๐Ÿš€ Features - persist open chat tabs across window close and relaunch (#2740) by @RaajeevChandran ## ๐Ÿ› Bug Fixes - make sure upgraders see the workspaces intro after the whats new modal (#2760) by @RaajeevChandran - show LAN discovered agents in the sidebar (#2758) by @RaajeevChandran - Use installed config and weightHigh9/14/2026
0.25.2## What's Changed - Add first-party n8n agent channel with secret-verified webhook ingress (#2748) by @tpae - Add Osaurus ID in Identity and fix Workspaces constant refresh (#2749) by @tpae - Show execution surface (sandbox VM vs native Mac) on tool approval cards (#2734) by @tpae - Computer Use audit: telemetry funnel, readiness gate, verified done, loop evals (#2731) by @tpae - Chat sidebar QoL: per-agent drafts, new-agent highlight, badge counts, subagent language, spell check (#2727) by @tpHigh9/13/2026
0.24.3## What's Changed - Advance the vmlx-swift pin to 1315fdcb (DeepseekV3 MoE f32 residual fix) (#2573) by @jjang-ai - Add Claude Code CLI integration (#2257) by @Yao920127 - Warn when switching models mid-conversation (#2538) by @jjang-ai - Repin vMLX for 90+ tok/s Ornith 35B decode (#2553) by @jjang-ai - Repin vMLX for scoped Ornith 35B MTP safety (#2549) by @jjang-ai - evals(community): OsaurusAI/Muse-Glimmer-30B-JANG\_6M on Apple M5 Max (#2544) by @jax-0n-git - Repin vMLX with Qwen 3.8 legacy-High8/31/2026
0.23.0## What's Changed - Attribute silent restarts: exit markers + a launch-time verdict file (#2401) by @jjang-ai - Reserve answer room on API requests so thinking cannot eat the whole max\_tokens (#2399) by @jjang-ai ## ๐Ÿš€ Features - added reasoning effort control for zai-glm models on mistral (#2412) by @RaajeevChandran - added projects support to group chats with shared instructions, knowledge and memory (#2337) by @RaajeevChandran ## ๐Ÿ› Bug Fixes - fix custom agents failing to spawn subagenHigh8/18/2026
0.22.17## What's Changed - Prove LFM2.5-2.6B (JANG\_6M + MXFP8) live and repin vMLX to the short-conv cache fixes (#2291) by @jjang-ai ## ๐Ÿš€ Features - add cmd+n shortcut option to start a new chat in the current window (#2296) by @RaajeevChandran - Add native WhatsApp agent channel (WhatsApp Web bridge) (#2290) by @tpae - Add "System" default theme card that follows macOS appearance (#2286) by @tpae ## ๐Ÿ› Bug Fixes - Extend release timeout for slow hosted runners (#2302) by @tpae - Fix custom ageHigh8/5/2026
0.22.10## What's Changed - Remove binary prefill proof artifacts (#2135) by @jjang-ai - Fix Bonsai DB recovery and SSD prefix reuse (#2153) by @jjang-ai - Fix TextEdit AppleScript state and completion handling (#2151) by @jjang-ai - Fix tool finalization and SSD cache warmup (#2146) by @jjang-ai ## ๐Ÿš€ Features - auto generate chat titles (#2144) by @RaajeevChandran - group consecutive thinking and tool activity into one expandable row (#2142) by @RaajeevChandran ## ๐Ÿ› Bug Fixes - fix settings windHigh7/24/2026
0.22.3## What's Changed ## ๐Ÿ› Bug Fixes - fixed find bar crash when scrolling to a match in a partially rendered message (#2020) by @RaajeevChandran - fixed mcp tool discovery dropping paginated tools/list results (#2017) by @RaajeevChandran - A request may only execute the tools it exposed (#2018) by @jjang-ai - Stop TTS from going silent after the output device changes (#2016) by @jjang-ai - Checkpoint: stop the gemma tool-loop, two crashes, and lying tok/s (#2015) by @jjang-ai ## ๐Ÿงฐ Maintenance High7/13/2026
0.21.1## What's Changed ## ๐Ÿ› Bug Fixes - Make built-in Default agent non-configurable; fix sub-agent toggle persistence (#1756) by @tpae - Fix chat history loss from non-idempotent schema migrations (#1755) by @tpae ## ๐Ÿงฐ Maintenance - Improve image models tab: inline Hugging Face link and download sizes (#1759) by @tpae - Rename Image Generation tab to Images (#1758) by @tpae - Fix native image generation echoing artifact JSON instead of confirming (#1757) by @tpae **Full Changelog**: https://gHigh6/29/2026
0.20.4## What's Changed - improved tool calls and prefill speed for small local models (#1577) by @RaajeevChandran - Add MCP Server Hub (#1559) by @mimeding - Allow manual provider models during connection tests (#1552) by @mimeding - Route sandbox plugin manifests through global proxy (#1549) by @mimeding - Route sandbox artifact downloads through global proxy (#1548) by @mimeding - Route relay tunnel traffic through global proxy (#1547) by @mimeding - Route remote agent invites through global proxyHigh6/18/2026
0.19.18## What's Changed - updated chinese translations (#1466) by @ftzahao - Document Gemma checkpoint memory safety contract (#1462) by @jjang-ai - Apply memory safety plan to model loads (#1461) by @jjang-ai - disabled whats new modal (#1458) by @RaajeevChandran - Expose memory safety status in cache stats (#1460) by @jjang-ai - Pin vMLX memory safety contract (#1456) by @jjang-ai - Expose native MTP runtime status (#1455) by @jjang-ai - Make RAM pressure advisory for local model loads (#1454) by @High6/11/2026
0.19.9## What's Changed - Fix Gemma4 reasoning, VL recognition, and tool schema rendering (#1342) by @jjang-ai - Pin Gemma4 nullable tool schema runtime (#1341) by @jjang-ai - Pin Gemma4 tool parser vMLX runtime (#1338) by @jjang-ai - Fix Gemma4 runtime CI guards (#1333) by @jjang-ai - [codex] Reject workbook targets in file\_write (#1332) by @mimeding - Fix Gemma4 chat image attachment routing (#1331) by @jjang-ai ## ๐Ÿ› Bug Fixes - fixed minor issue with chat template (#1343) by @tpae - fixed mainHigh6/5/2026
0.19.2## What's Changed - chat polish with smoother streaming, stable tool call animations and scroll cache fixes (#1283) by @RaajeevChandran - [codex] Harden StatsPack schema metadata (#1280) by @mimeding - [codex] Harden PPTX ingress metadata (#1279) by @mimeding - [codex] Harden generated XLSX cell output (#1278) by @mimeding - [codex] Preview XLSX workbooks through file\_read (#1277) by @mimeding - [codex] Harden MCP OAuth metadata redirects (#1276) by @mimeding - [codex] Contain skill file pathsHigh5/28/2026
0.18.36## What's Changed - Fix capability search prompt arguments (#1208) by @mimeding - Accept Responses message input without type (#1206) by @mimeding - Narrow trivial local chat tool schema (#1205) by @mimeding - Stabilize storage-path CI gate (#1204) by @mimeding ## ๐Ÿš€ Features - Share Themes and import themes with deep link (#1202) by @tpae ## ๐Ÿงฐ Maintenance - onboarding copy changes (#1210) by @tpae - reduce avatar size on empty chat view (#1209) by @tpae - sped up sandbox launch (#1207) byHigh5/22/2026
0.18.19## What's Changed - Fix Gemini tool schema parameters (#1088) by @mimeding ## ๐Ÿ› Bug Fixes - fixed issue with reverting changes in capabilities manager (#1100) by @tpae - prevent switching active theme when saving a theme edit (#1096) by @RaajeevChandran - improved theme edit sync in chat (#1095) by @RaajeevChandran **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.18.18...0.18.19 High5/15/2026
0.18.13## What's Changed - custom avatar support for agents (#1051) by @RaajeevChandran ## ๐Ÿงฐ Maintenance - moved sandbox secrets to agent details view (#1053) by @tpae - improve accessibility in light themes (#1052) by @RaajeevChandran **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.18.12...0.18.13 High5/8/2026
0.18.9## What's Changed - auto speak tts in chat (#1026) by @RaajeevChandran ## ๐Ÿš€ Features - added insights details tab (#1034) by @tpae ## ๐Ÿ› Bug Fixes - chat history \& schedule creation improvements (#1032) by @RaajeevChandran - hide folder tool guidance in auto selection mode (#1028) by @RaajeevChandran ## ๐Ÿงฐ Maintenance - reverted system prompt changes (#1029) by @RaajeevChandran **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.18.8...0.18.9 High5/5/2026
0.18.2## What's Changed - [codex] Gate local thinking controls and SQLCipher import (#954) by @mimeding - improved model manager (#961) by @RaajeevChandran ## ๐Ÿ› Bug Fixes - prevent cold loading large core MLX models on launch for memory distillation (#970) by @RaajeevChandran ## ๐Ÿงฐ Maintenance - minor changes (#972) by @RaajeevChandran - added core model fallback if missing (#971) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.18.1...0.18.2 High4/29/2026
0.17.6## What's Changed - feat(model-picker): add Performance filter for 'Runs Well' / 'Hide Too Large' (#946) by @jjang-ai - fix(speech): adopt FluidAudio 0.14 decoderState parameter (#945) by @mimeding - fix(security): harden import/export trust boundaries for attachments and shared artifacts (#925) by @mimeding - fix(provider): send max\_tokens to non-OpenAI-reasoning endpoints (#928) by @mimeding ## ๐Ÿš€ Features - added OpenAI log in with Codex (subscription) (#947) by @tpae ## ๐Ÿงฐ Maintenance High4/24/2026
0.17.1## What's Changed * No changes **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.17.0...0.17.1 High4/21/2026
0.17.0## What's Changed - added vision support flag in http handler (#904) by @RaajeevChandran ## ๐Ÿ› Bug Fixes - fix failing tests (#907) by @tpae ## ๐Ÿงฐ Maintenance - clean up onboarding views (#909) by @tpae - fixed thinking block glitch (#908) by @RaajeevChandran - clarification UI updates (#906) by @tpae - address gaps in host API + background tasks management (#905) by @tpae - Deprecate Work Mode (migrate into single Chat / Agent system) (#893) by @tpae **Full Changelog**: https://github.comHigh4/21/2026
0.16.16## What's Changed - feat: Qwen 3.6 + JANGTQ support + Settings crash + symlink loading fixes (#878) by @jjang-ai - Fix pairing approval Enter shortcut ignoring persistent choice (#876) by @shaun0927 - chore: lint (#874) by @ritave - feat: complete German localization (Phase 3-6) (#837) by @HolliOnRoad ## ๐Ÿ› Bug Fixes - accept model param in /api/show to match ollama API spec (#882) by @RaajeevChandran - fix: show user-configured remote providers in model picker (#883) by @jjang-ai - partitionHigh4/17/2026
0.16.15## What's Changed - feat: complete German localization (Phase 3-6) (#837) by @HolliOnRoad - [codex] Add Simplified Chinese localization (#857) by @jiajun-dev - feat: allow bonjour agents to pair automatically and use relay (#870) by @ritave ## ๐Ÿ› Bug Fixes - Template-driven reasoning detection + streaming/retroactive thinking fixes (#867) by @jjang-ai **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.16.14...0.16.15 High4/16/2026
0.16.14## What's Changed - removed unused theme properties \& improved code block customization/readability (#864) by @RaajeevChandran ## ๐Ÿ› Bug Fixes - Fix stuck loading, #856 stats leak, #858 error classifier, #860 NIO crash + bump vmlx-swift-lm (#863) by @jjang-ai **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.16.13...0.16.14 High4/15/2026
0.16.13## What's Changed - feature: strengthen Work Mode completion contract (#832) by @mimeding ## ๐Ÿงฐ Maintenance - fixed artifact sharing on plugin (#859) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.16.12...0.16.13 High4/15/2026
0.16.12## What's Changed ## ๐Ÿงฐ Maintenance - bump vmlx version and added repair models function (#855) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.16.11...0.16.12 High4/14/2026
0.16.11## What's Changed ## ๐Ÿ› Bug Fixes - sandbox fixes and other updates (#853) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.16.10...0.16.11 High4/14/2026
0.16.10## What's Changed - VMLX migration (#849) by @tpae - Fix tool calls from OpenResponses (#850) by @ritave - optimize TTFT for local models (#847) by @RaajeevChandran - inline data visualization (#829) by @RaajeevChandran ## ๐Ÿ› Bug Fixes - added missing jang config (#845) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.16.9...0.16.10 Medium4/13/2026
0.16.9## What's Changed - slash commands (#822) by @RaajeevChandran ## ๐Ÿ› Bug Fixes - fix Stage Manager window behavior and model change window activation (#824) by @RaajeevChandran - fix: auto-enable MCP and sandbox plugin tools on first registration (#827) by @jjang-ai ## ๐Ÿงฐ Maintenance - updated acknowledgements (#826) by @RaajeevChandran **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.16.8...0.16.9 High4/9/2026
0.16.8## What's Changed - fix: hide placeholder during IME composition to prevent text overlap (#819) by @SliverKeigo - Performance \& Visual improvements (#820) by @RaajeevChandran - feat: German localization (821 strings) (#785) by @HolliOnRoad **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.16.7...0.16.8 Medium4/9/2026
0.16.7## What's Changed - optimize tiled background image rendering (#812) by @RaajeevChandran - feat: upgrade OpenAI api to use new /v1/responses endpoint (#815) by @ritave ## ๐Ÿ› Bug Fixes - updated preflight (auto) tool selection to be more reliable (#818) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.16.6...0.16.7 High4/8/2026
0.16.6## What's Changed ## ๐Ÿš€ Features - added turboquant configuration (#810) by @tpae ## ๐Ÿ› Bug Fixes - fixed background image for work view (#811) by @tpae ## ๐Ÿงฐ Maintenance - context management cleanup (#807) by @tpae - chore: bump mlx-swift-lm to 7d9a6ab (#809) by @jjang-ai **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.16.5...0.16.6 Medium4/8/2026
0.16.5## What's Changed - fix: bonjour agent calls now run tools on the server (#767) by @ritave ## ๐Ÿงฐ Maintenance - cleaned up model manager (#805) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.16.4...0.16.5 Medium4/7/2026
0.16.4## What's Changed ## ๐Ÿ› Bug Fixes - fix: crash on second generation with Gemma 4 hybrid KV cache (#799) by @jjang-ai - fix app window focus and activation on launch (#794) by @RaajeevChandran - fix: VLM classification false positives and image persistence across model switches (#795) by @jjang-ai ## ๐Ÿงฐ Maintenance - added osaurus curated models (#800) by @tpae - bumped versions on mlx lm library (#797) by @tpae - added metal gate to prevent crashes (#796) by @tpae **Full Changelog**: https:Medium4/7/2026
0.16.3## What's Changed ## ๐Ÿงฐ Maintenance - added stats and fixed minor crash issues (#792) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.16.2...0.16.3 Medium4/6/2026
0.16.2## What's Changed ## ๐Ÿ› Bug Fixes - fix: built-in tool discovery and chat view crash (#791) by @jjang-ai **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.16.1...0.16.2 Medium4/6/2026
0.16.1## What's Changed ## ๐Ÿ› Bug Fixes - updated skills view to be more deterministic with calls (#787) by @tpae ## ๐Ÿงฐ Maintenance - fix crashes, extend gemma support (#786) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.16.0...0.16.1 Medium4/6/2026
0.16.0## What's Changed - added factory reset in settings (#780) by @RaajeevChandran ## ๐Ÿš€ Features - manual tool selection (#771) by @tpae ## ๐Ÿ› Bug Fixes - address known swiftUI bug for skills import button and fixes few other crashes (#784) by @tpae - dynamic storage monitoring for custom model directories (#781) by @RaajeevChandran - fix: PDF attachment 'Document appears to be empty' error (#779) by @jjang-ai - fixed issue with model downloader (#778) by @tpae - preview and save image in artiMedium4/5/2026
0.15.18## What's Changed - automatic detection of clipboard content in chat window (#764) by @RaajeevChandran ## ๐Ÿš€ Features - updated autonomous execution mode - auto plugin creation flow (#765) by @tpae ## ๐Ÿ› Bug Fixes - fix UI Freeze by making skill related I/O and processing asynchronous (#763) by @RaajeevChandran ## ๐Ÿงฐ Maintenance - minor UI improvements in chat view (#766) by @RaajeevChandran **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.16...0.15.17 Medium4/2/2026
0.15.16## What's Changed - chore: added linting on pre-push hook using lefthook (#761) by @ritave - feat: add bonjour broadcast and discovery for agents (#683) by @ritave - GPU optimizations (#759) by @RaajeevChandran ## ๐Ÿ› Bug Fixes - updated relay to refresh after load (#758) by @tpae ## ๐Ÿงฐ Maintenance - fix team id (#762) by @tpae - fixed embedding / skills related potential crashes, extended unit tests (#760) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.15..Medium4/1/2026
0.15.15## What's Changed ## ๐Ÿงฐ Maintenance - fixed issues with skills import (#757) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.14...0.15.15 Medium3/31/2026
0.15.14## What's Changed ## ๐Ÿงฐ Maintenance - release process updates (#756) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.13...0.15.14 Medium3/31/2026
0.15.13## What's Changed ## ๐Ÿงฐ Maintenance - fixed package versioning (#755) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.12...0.15.13 Medium3/30/2026
0.15.12## What's Changed ## ๐Ÿงฐ Maintenance - fixed crash with invalid plugins (#753) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.11...0.15.12 Medium3/29/2026
0.15.11## What's Changed ## ๐Ÿงฐ Maintenance - fixed crash; added additional guards for metal paths, removed async prefix warming (#752) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.10...0.15.11 Medium3/29/2026
0.15.10## What's Changed ## ๐Ÿงฐ Maintenance - improve concurrency with agent handling (#751) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.9...0.15.10 Medium3/29/2026
0.15.9## What's Changed - updated FluidAudio to latest version (#750) by @RaajeevChandran ## ๐Ÿงฐ Maintenance - Plugin updates - major improvements to Telegram plugin (#749) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.8...0.15.9 Medium3/29/2026
0.15.8## What's Changed - batch embedding for context index rebuilds (#748) by @RaajeevChandran - feat: add Disable Tools toggle to chat configuration (#735) by @vernonstinebaker ## ๐Ÿ› Bug Fixes - updated rag index init to be serial to prevent race conditions (#747) by @tpae - fix race condition (#746) by @RaajeevChandran **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.7...0.15.8 Medium3/29/2026
0.15.7## What's Changed - several optimizations for bringing down GPU usage (#738) by @RaajeevChandran - perf: two-phase prefill for hybrid SSM/transformer models (Qwen3.5) (#737) by @vernonstinebaker - perf: tiered KV cache store, O(1) streaming decode, eval fix (#736) by @vernonstinebaker - perf: memoize preflight capability search results per session (#734) by @vernonstinebaker - perf: sort alwaysLoadedSpecs() for deterministic tool ordering (#733) by @vernonstinebaker - chore: bump swift-jinja 2.Medium3/27/2026
0.15.6## What's Changed - fix(tools): delegate MLX tool-call parsing to upstream ToolCallProcessor (#714) by @vernonstinebaker - chore: fix Swift 6 / Xcode 26.4 strict concurrency build errors (#727) by @vernonstinebaker ## ๐Ÿงฐ Maintenance - updated tools dev and local plugin access (#729) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.5...0.15.6 Medium3/26/2026
0.15.5## What's Changed - perf(PrefixCache): lazy load prefix cache in background to drastically unblock TTFT (#717) by @vernonstinebaker - improved scroll performance in model picker (#724) by @RaajeevChandran - fix(TypingIndicator): fix shimmer overflow into "Processing promptโ€ฆ" text (#725) by @vernonstinebaker ## ๐Ÿงฐ Maintenance - resolve sym link for tool cli build (#726) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.4...0.15.5 Medium3/26/2026
0.15.4## What's Changed ## ๐Ÿงฐ Maintenance - improve plugin dev experience (#723) by @tpae - Host API updates - added preflight and artifact handler (#722) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.3...0.15.4 Medium3/26/2026
0.15.3## What's Changed ## ๐Ÿ› Bug Fixes - fixed default tool indexing issues (#721) by @tpae - fixed issue with missing remote providers for work and host api (#720) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.2...0.15.3 Medium3/25/2026
0.15.2## What's Changed - Fix ChatEngine remote services isolation, bookmark fallback, and config trimming (#712) by @vernonstinebaker - Rewrite StreamAccumulator as pull-based AsyncSequence for reliable local inference (#711) by @vernonstinebaker - Add prefill progress bar and auto-expand thinking block during streaming (#710) by @vernonstinebaker - added manual transcription stop mode (#701) by @RaajeevChandran ## ๐Ÿ› Bug Fixes - fixed 400 issue with xAI API with missing tool params (#718) by @tpaMedium3/25/2026
0.15.1## What's Changed - feat: Add MCPB bundle loading CLI command (#706) by @drernie - Add InferenceProgressManager for tracking MLX prefill/generation progress (#709) by @vernonstinebaker - Fix isValidCache to accept hybrid (Mamba+KVCacheSimple) model caches (#708) by @vernonstinebaker - fix: plugins didn't load secrets for agents properly (#713) by @ritave **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.15.0...0.15.1 Medium3/24/2026
0.15.0## What's Changed - ensure menu bar panel and windows are visible over full screen apps (#702) by @RaajeevChandran - smart toggle logic for global chat hotkey (#703) by @RaajeevChandran ## ๐Ÿš€ Features - Improved Context Management System - new concept: methods (#704) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.14.29...0.14.30 Medium3/24/2026
0.14.29## What's Changed - batch process voice input instead of real time (#699) by @RaajeevChandran - UI/UX improvements (#698) by @RaajeevChandran - open settings page with previously selected sidebar item (#697) by @RaajeevChandran ## ๐Ÿงฐ Maintenance - use latest version of vectura kit (#700) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.14.28...0.14.29 Low3/22/2026
0.14.28## What's Changed - improved UX for thinking toggle (#694) by @RaajeevChandran ## ๐Ÿ› Bug Fixes - fixed crash with gpu release (#693) by @tpae ## ๐Ÿงฐ Maintenance - added additional guards for streaming (#695) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.14.27...0.14.28 Low3/21/2026
0.14.27## What's Changed - eliminate idle GPU/CPU pressure in FloatingInputCard (#688) by @RaajeevChandran ## ๐Ÿงฐ Maintenance - added cache guard to prevent crashes (#691) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.14.26...0.14.27 Low3/20/2026
0.14.26## What's Changed ## ๐Ÿงฐ Maintenance - lazy load KV cache for better user experience (#687) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.14.25...0.14.26 Low3/20/2026
0.14.25## What's Changed - fix: Telegram plugin not working after restart (#685) by @ritave ## ๐Ÿš€ Features - work mode improvements (#684) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.14.24...0.14.25 Low3/20/2026
0.14.24## What's Changed ## ๐Ÿ› Bug Fixes - fixed authentication missing from remote providers (#680) by @tpae ## ๐Ÿงฐ Maintenance - updated code structure for better readability (#679) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.14.23...0.14.24 Low3/20/2026
0.14.23## What's Changed ## ๐Ÿงฐ Maintenance - local model prefill cache optimizations (#677) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.14.22...0.14.23 Low3/19/2026
0.14.22## What's Changed - ensure that new chat windows use the previously active window size (#646) by @RaajeevChandran - add support for cron expression scheduling (#673) by @RaajeevChandran ## ๐Ÿงฐ Maintenance - polished up visual look on chat / work views (#674) by @tpae - updated minor documentation changes (#670) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.14.21...0.14.22 Low3/19/2026
0.14.21## What's Changed ## ๐Ÿš€ Features - GPU Pinning for efficient memory usage and speed optimization (#669) by @tpae ## ๐Ÿงฐ Maintenance - fixed lfm2 tool calling issues (#668) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.14.20...0.14.21 Low3/18/2026
0.14.20## What's Changed ## ๐Ÿ› Bug Fixes - show all agents for relay toggle (#667) by @tpae - fixed double dispatch for transcription service (#666) by @tpae ## ๐Ÿงฐ Maintenance - expose sandbox secrets to tools (#665) by @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.14.19...0.14.20 Low3/18/2026
0.14.19## What's Changed ## ๐Ÿงฐ Maintenance - Fixed sandbox plugin related issues (#663) by @tpae - clean up UI in Work View (#661) by @tpae - added minor capabilities warning (#659) by @tpae ## New Contributors @github-actions[bot] and @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.14.18...0.14.19 Low3/18/2026
0.14.18## What's Changed ## ๐Ÿ› Bug Fixes - fixed issue with warm up crash and updated sandbox tools loading (#658) by @tpae - fixed issue with sandbox secrets injection (#657) by @tpae ## New Contributors @github-actions[bot] and @tpae **Full Changelog**: https://github.com/osaurus-ai/osaurus/compare/0.14.17...v0.14.18 Low3/18/2026
0.14.17## What's Changed - fix sandbox plugin modal overflow - align reasoning model params with OpenAI spec \& add 'minimal' effort support - prioritize custom auth headers \& prevent automatic overwrite in remote providers ## ๐Ÿš€ Features - Smart KV Caching - improved local model performance significantly Low3/18/2026
0.14.16## What's Changed - UI/UX enhancements - implemented window frame persistence for chat \& settings windows - fix: tapping active mode segment incorrectly switches to opposite mode ## ๐Ÿงฐ Maintenance - renamed model names for clarity Low3/15/2026

Dependencies & License Audit

Loading dependencies...

Similar Packages

voicemodeNatural (2-way) voice conversations with Claude Codev9.0.0
studioOpen-source control plane for your AI agents. Connect tools, hire agents, track every token and dollarv4.519.0
mcp-ts-coreAgent-native TypeScript framework for building MCP servers. Build tools, not infrastructure.v0.13.16
jdocmunch-mcpThe leading, most token-efficient MCP server for documentation exploration and retrieval via structured section indexingv1.150.1
mcp-searxngMCP Server for SearXNGv2.7.0

More in MCP Servers

litellmPython SDK, Proxy Server (AI Gateway) to call 100+ LLM APIs in OpenAI (or native) format, with cost tracking, guardrails, loadbalancing and logging. [Bedrock, Azure, OpenAI, VertexAI, Cohere, Anthropi
claude-plugins-officialOfficial, Anthropic-managed directory of high quality Claude Code Plugins.
casdoorAn open-source Agent-first Identity and Access Management (IAM) /LLM MCP & agent gateway and auth server with web UI supporting OpenClaw, MCP, OAuth, OIDC, SAML, CAS, LDAP, SCIM, WebAuthn, TOTP, MFA,
hyperframesWrite HTML. Render video. Built for agents.