Maka
SHA-256Maka is a local-first AI agent workspace that logs every model message, tool call, and permission decision in an append-only Runtime Event Log. Sessions, UI, and recovery are projections of that log. Choose your own model connection—cloud, local, or gateway—while all data stays on your machine by default.
Smart Download
Download Download Version
v0.1.11 · 246 MB
Local-first AI agent workspace with a recoverable append-only log for every model message and tool call.
Core Features
- Local-first: sessions, settings, and run records stay on your machine by default
- Append-only Runtime Event Log: all model messages, tool calls, results, and termination events are recorded
- Multiple entry points: Desktop (Electron + React), TUI, CLI, and Eval all share the same Runtime Host
- Controlled permissions: file writes, Shell, and dangerous tools go through a permission engine
- Context optimization: Tool Result pruning and LLM Compaction change inference input without deleting evidence
What It Can't Do
- •Currently only macOS Apple Silicon (arm64) is supported; Windows is an unsigned preview, and Intel Mac / Linux are not supported yet. Maka does not bundle any model account—you must configure your own model connection. The project is under active development, so data formats, CLI commands, and features may change. On Windows, if SmartScreen warns about the publisher, do not bypass it unless the downloaded SHA-256 matches the checksum published with the release.
Use Cases
- Developers who want a private, local AI agent for coding, project inspection, and command execution
- Security-sensitive environments that require auditable records of every AI decision and tool action
Detailed Introduction
Maka is an Apache Incubating local-first AI agent workspace designed for real work. It goes beyond simple chat by letting agents inspect projects, execute tools, and produce artifacts—all under controlled permissions. The core differentiator is the Runtime Event Log: every model message, tool call, result, and termination event is recorded as an append-only log. Sessions, UI, model context, and recovery are all projections of that log, making execution facts recoverable and auditable. Compared to cloud-hosted AI assistants, Maka is privacy-first by default: your sessions, settings, and run records stay on your machine, and you choose the model connection (cloud API, local model, or compatible gateway). It offers multiple surfaces: a desktop app (Electron + React), a terminal TUI, a non-interactive CLI, and a benchmark evaluation framework. The architecture separates execution authority (Runtime Host) from experiment semantics (Eval), ensuring reproducible and safe agent behavior. While still under active development, it already supports streaming, tool schema validation, permission policies, watchdog timers, branching sessions, and history compaction.
Tags
Getting Started
Download installer
Click the button above to download the installer for your system
Install the software
Open the downloaded dmg file, then drag the app to Applications
Download the macOS arm64 DMG from GitHub Releases and open it
Drag Maka into your Applications folder
Install ripgrep with brew install ripgrep to enable the Grep tool
- Download the macOS arm64 DMG from GitHub Releases and open it
- Drag Maka into your Applications folder
- Install ripgrep with brew install ripgrep to enable the Grep tool
SHA-256 checksum verified
Checksum extracted from GitHub official Release page
SHA256 Checksum
854e7db056e163dcb649db9c5d7164fd36bf596fc8d095d53f1e9798028a4af2This checksum is extracted from the GitHub Release page. Verify file integrity after download.
All SHA-256 checksums on this platform are extracted from the project's official GitHub Release page, without any modification. You can independently verify them on the GitHub Releases page.
Open Source Transparency
View GitHub SourceUninstall Info
macOS: drag Maka.app to Trash; to remove all data, delete the Electron userData folder (e.g., ~/Library/Application Support/Maka). Windows: uninstall via Settings, then remove the remaining C:\Users\<user>\AppData\Roaming\Maka directory.
No Extra Dependencies
Ready to use after download. No additional runtime required.
Similar Projects
ragflow
RAGFlow is a leading open-source RAG engine that fuses deep document understanding with agentic workflows, offering precise, citation-grounded answers. Unlike LangChain's DIY approach, RAGFlow provides a turnkey, enterprise-ready platform with a visual interface and template-based chunking, reducing hallucinations and setup complexity.
unsloth
Unsloth is the first desktop app to run and train AI models locally. It supports a wide range of models including Qwen3.8, Kimi K3, MiniMax-H3, Gemma 4, DeepSeek-V4, and FLUX. Unlike most local model tools that focus only on inference, Unsloth adds a full training stack: fine-tune LLMs, diffusion, TTS, and embedding models 2× faster with 70% less VRAM. It also supports LoRA, QLoRA, RL, GRPO, DPO, and FP8. The app includes built-in Agents & Tools integration (Claude Code, Codex, MCP), private search and RAG, image/video diffusion, and audio support. With hardware support for CPU, NVIDIA, AMD, Intel, macOS, and multi-GPU setups, plus remote access via Cloudflare, Unsloth is a comprehensive local AI studio. It can export to GGUF, NVFP4, FP8 and offers an OpenAI-compatible API for easy deployment.
LocalAI
LocalAI is the open-source AI engine to run any model - LLMs, vision, voice, image, video - on any hardware. No GPU required. Drop-in API compatibility with OpenAI, Anthropic, and ElevenLabs.