# liliMozi/openhanako

A personal AI agent with memory, personality, and autonomy.

Repository: https://github.com/liliMozi/openhanako
Canonical: https://ross.abutalabs.com/products/openhanako
Language: TypeScript
License: Apache-2.0
License Family: permissive
Last push: 2026-08-22T02:03:20+00:00

## Health v2 (maintenance only)
Score: 77/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 98, release rhythm 86, longevity 12
- inputs: {"age_days": 171, "days_push": 12, "days_rel": 12, "gap_med": 0, "n_releases_24m": 124}
- flags: young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 6362, forks 551 (observed 2026-08-28T04:09:42.837247+00:00)

## What it is
HanaAgent is a cross-platform desktop application providing a personal AI agent with persistent memory, customizable personality, and autonomous/proactive behavior, aimed at everyday computer users rather than only developers. It bundles a full GUI, multi-agent collaboration, file/terminal/web tools, a community skills ecosystem, a plugin system, OS-level sandboxing, and bridges to messaging platforms like Telegram, WeChat, QQ, and Feishu.

## Use cases
- personal AI assistant with long-term memory
- AI agent that can control my computer and run commands
- AI assistant with a customizable personality
- run multiple AI agents that collaborate on my desktop
- chat with my desktop AI agent from Telegram or WeChat
- automate daily office tasks with scheduled AI jobs
- self-hosted AI agent with a graphical interface
- AI assistant that learns new skills on its own

## When to choose
- You want a GUI-first AI agent usable by non-programmers, unlike CLI coding agents such as Claude Code or Codex
- You need per-agent memory, personality files, and isolated agent folders that are easy to back up or share as character cards
- You want an agent that can read/write files, run terminal sessions, browse the web, and take screenshots for everyday office work
- You want the same agent reachable from Telegram, WeChat, QQ, or Feishu plus a mobile PWA over LAN
- You value multi-agent coordination, cron-style scheduled tasks, and a sandboxed permission model (Seatbelt/Bubblewrap/restricted token)

## When to avoid
- You need a lightweight CLI coding agent tightly integrated into a terminal or IDE development workflow
- You want a library or SDK to embed agent capabilities into your own product rather than a standalone end-user application
- You require enterprise-grade, audited security; the community skills/plugins and sandbox model are geared toward personal use
- You need fine-grained programmatic control over prompts, context windows, or model routing that raw agent SDKs provide

## Facets
- artifact type: application
- maturity: active
- function: agent-framework, chatbot, chat-interface, workflow-automation, scheduling, plugin-system, developer-tools
- domain: artificial-intelligence, chatbots, large-language-models, developer-tools, self-hosted
- platform: windows, cross-platform, cli, self-hosted
- tags: personal-ai-assistant, persistent-memory, agent-personality, multi-agent, skills-system, character-cards, os-level-sandbox, cron-tasks, proactive-agent, messaging-bridge, desktop-gui, pwa, plugin-ecosystem, ai-agents, automation, macos, linux, desktop, mobile, telegram, wechat

## Member repositories
- liliMozi/openhanako (main) score 77

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:42.837247+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:46:04.676940+00:00, confidence not recorded.
  - readme: https://github.com/liliMozi/openhanako (fetched 2026-08-28T04:09:42.837247+00:00, sha 47805754109f)
- Data as of 2026-08-30T08:39:29.467469+00:00.