Back to directory

SUZENT

An open-source, local-first personal AI agent whose identity, memory and workspace stay under the user's control, usable with GPT, Claude, Gemini, DeepSeek or local models without losing continuity when switching between them.

SUZENT

SUZENT stores its memory as append-only markdown files on the user's own disk rather than in a hosted database, runs its tool calls through permission modes the user defines, and isolates execution inside Docker workspaces the user owns. It can research, write, code, pursue longer-running goals, run scheduled work, and connect to a user's devices, meeting them through Telegram, Slack, Discord, Feishu or WeChat within the boundaries the user sets, while the underlying model provider can be swapped without resetting the agent's accumulated memory and skills.

Installation is a single setup script for Windows, macOS or Linux that installs a Python backend and a command-line tool, needing only Git as a prerequisite; a mirror mode speeds up the same installation using alternative package mirrors. Updates are applied as one matched set of backend source, locked dependencies and desktop application, performed by a standalone updater running outside the active virtual environment.

It is written in Python and released under the Apache-2.0 license.

More in Agent frameworks

See all 902
  • Recurse

    recurse.run

    A serverless harness that lets a coding agent design, test, and deploy its own specialist sub-agents as tools, MCP servers, or bots, defined by a schema-validated manifest.

  • ProofForge

    shulgin.is-a.dev

    An AI-agent pipeline that decomposes a mathematics problem, proves each piece, and formalizes the result in Lean 4 and Mathlib until the Lean kernel accepts it, with six of its proofs already merged into Google DeepMind's formal-conjectures repository.

  • local-coder

    github.com

    Builds a local coding-agent team on top of Ollama and OpenCode by detecting the machine's hardware, recommending models for seven specialist roles, and testing whether each assigned model can actually use tools, edit a repository and pass verification.