Your Autonomous Local AI Assistant & Pair Programmer.
High-performance, memory-isolated desktop daemon built in pure Rust. Native tool calling, terminal execution, and multi-model routing with 100% private execution.
priora-installer
PowerShell (Run as User or Administrator)
1-Click Install
irm https://prioraagent.com/install.ps1 | iex
Terminal (Bash or Zsh)
1-Click Install
curl -fsSL https://prioraagent.com/install | sh
Auto-detects x86_64 and ARM64 architectures
Zero-telemetry & 100% local binary
Engineered for Precision & Sovereignty
Built from scratch with zero Electron bloat. Fast local IPC, AST parsing, and sandboxed execution.
Native Rust Daemon
Compiles to a single lightweight ~15MB binary. Starts in under 20ms with near-zero idle RAM consumption.
Memory Footprint
< 25 MB RAM
Multi-Model Routing
Seamlessly orchestrate local Ollama models (Qwen, Llama, DeepSeek) or fallback to frontier models via OpenRouter.
Routing Mode
Local-first & Hybrid
Workspace Trust Matrix
Granular permission controls. Run in read-only analysis, guided mode, or full autopilot with automated Git checkpoints.
Safety Model
Git rollback ready
Architecture Overview
Obsidian Velocity Engine
Port 31415
SSE Streams
Headless IPC Daemon: Serves the Web Studio at
http://127.0.0.1:31415 with real-time SSE progress events.
Autonomous Living Spec: Auto-generates test gates, implementation todos, and AST-driven codebase maps.
Git Checkpoint Snapshots: Automatic micro-checkpoints before every agent run, allowing instant 1-click rollback.
CLI Companion: Seamless terminal command interface (
dp or priora) directly integrated with your shell.