mirror of
https://github.com/memohai/Memoh.git
synced 2026-04-25 07:00:48 +09:00
sheepbox8646
2 lines
8.3 KiB
JavaScript
2 lines
8.3 KiB
JavaScript
import{_ as t,o as a,c as o,ag as s}from"./chunks/framework.CAXxHpAX.js";const g=JSON.parse('{"title":"About Memoh","description":"","frontmatter":{},"headers":[],"relativePath":"about.md","filePath":"about.md","lastUpdated":1776332711000}'),i={name:"about.md"};function n(r,e,l,d,h,c){return a(),o("div",null,[...e[0]||(e[0]=[s('<h1 id="about-memoh" tabindex="-1">About Memoh <a class="header-anchor" href="#about-memoh" aria-label="Permalink to "About Memoh""></a></h1><h2 id="what-is-memoh" tabindex="-1">What Is Memoh? <a class="header-anchor" href="#what-is-memoh" aria-label="Permalink to "What Is Memoh?""></a></h2><p>Memoh is a multi-member, structured long-memory, containerized AI agent platform. You can create multiple AI bots, give each bot its own isolated workspace and long-term memory, and interact with them through Telegram, Discord, Lark (Feishu), QQ, Matrix, Misskey, DingTalk, WeCom, WeChat, WeChat Official Account, Email, or the built-in Web UI.</p><p>Every bot has its own execution environment, tools, memory configuration, and channel integrations. In practice, that means each bot behaves more like its own computer-backed agent than a shared chat preset.</p><h2 id="what-makes-memoh-different" tabindex="-1">What Makes Memoh Different <a class="header-anchor" href="#what-makes-memoh-different" aria-label="Permalink to "What Makes Memoh Different""></a></h2><h3 id="multi-bot-and-multi-user" tabindex="-1">Multi-Bot And Multi-User <a class="header-anchor" href="#multi-bot-and-multi-user" aria-label="Permalink to "Multi-Bot And Multi-User""></a></h3><p>Memoh is built for real sharing and real separation at the same time:</p><ul><li>create multiple bots for different roles or people</li><li>let humans and bots interact in private chats, groups, or delegated workflows</li><li>distinguish individual users in shared conversations</li><li>bind identities across channels so the same person can be recognized consistently</li></ul><h3 id="containerized-workspaces" tabindex="-1">Containerized Workspaces <a class="header-anchor" href="#containerized-workspaces" aria-label="Permalink to "Containerized Workspaces""></a></h3><p>Each bot runs in its own isolated container workspace with a separate filesystem and network boundary. Bots can read and write files, run commands, and use tools inside that workspace without interfering with other bots.</p><h3 id="long-term-memory-and-context-management" tabindex="-1">Long-Term Memory And Context Management <a class="header-anchor" href="#long-term-memory-and-context-management" aria-label="Permalink to "Long-Term Memory And Context Management""></a></h3><p>Memoh separates two different problems:</p><ul><li><strong>Long-term memory</strong> stores durable facts and recalls them across conversations through memory providers</li><li><strong>Session context compaction</strong> reduces the prompt size of an active session when the current conversation gets too large</li></ul><p>This distinction is important: context compaction changes the active session window, while memory compaction rewrites stored memory entries.</p><h3 id="sessions-and-discuss-mode" tabindex="-1">Sessions And Discuss Mode <a class="header-anchor" href="#sessions-and-discuss-mode" aria-label="Permalink to "Sessions And Discuss Mode""></a></h3><p>Each bot maintains independent <strong>sessions</strong> that preserve context. Memoh currently uses five session types:</p><ul><li><strong>Chat</strong> — regular user-facing conversations</li><li><strong>Discuss</strong> — deliberative sessions where the bot can think through work and decide what to send outward</li><li><strong>Heartbeat</strong> — periodic autonomous sessions</li><li><strong>Schedule</strong> — cron-triggered task sessions</li><li><strong>Subagent</strong> — delegated task sessions</li></ul><p>You can start or route sessions with slash commands such as <code>/new</code>, and the Web UI exposes a session status panel with metrics like context usage, cache hit rate, and used skills.</p><h3 id="broad-channel-coverage" tabindex="-1">Broad Channel Coverage <a class="header-anchor" href="#broad-channel-coverage" aria-label="Permalink to "Broad Channel Coverage""></a></h3><p>Memoh uses a unified channel adapter system so one bot can be reachable from many places at once.</p><p>Current user-facing integrations include:</p><ul><li><strong>Telegram</strong></li><li><strong>Discord</strong></li><li><strong>Lark (Feishu)</strong></li><li><strong>QQ</strong></li><li><strong>Matrix</strong></li><li><strong>Misskey</strong></li><li><strong>DingTalk</strong></li><li><strong>WeCom</strong></li><li><strong>WeChat</strong></li><li><strong>WeChat Official Account</strong></li><li><strong>Email</strong></li><li><strong>Web</strong></li></ul><p>Memoh also distinguishes between the personal <strong>WeChat</strong> QR-login integration and the webhook-based <strong>WeChat Official Account</strong> integration.</p><h3 id="tools-skills-mcp-and-supermarket" tabindex="-1">Tools, Skills, MCP, And Supermarket <a class="header-anchor" href="#tools-skills-mcp-and-supermarket" aria-label="Permalink to "Tools, Skills, MCP, And Supermarket""></a></h3><p>Bots can use a rich set of built-in capabilities, including:</p><ul><li>web search and web fetch</li><li>browser automation</li><li>file editing and command execution inside the bot workspace</li><li>memory search and management</li><li>messaging, email, and TTS</li><li>subagents for delegated work</li><li><strong>skills</strong> for reusable behavior modules</li><li><strong>MCP</strong> connections for external tool servers</li><li><strong>Supermarket</strong> for curated skill and MCP template installation</li></ul><h3 id="providers-and-models" tabindex="-1">Providers And Models <a class="header-anchor" href="#providers-and-models" aria-label="Permalink to "Providers And Models""></a></h3><p>Memoh supports multiple provider client types, including:</p><ul><li>OpenAI-compatible chat completions</li><li>OpenAI Responses API</li><li>Anthropic Messages</li><li>Google Generative AI</li><li>OpenAI Codex</li><li>GitHub Copilot</li><li>Edge Speech / TTS</li></ul><p>Models are also separated by role:</p><ul><li><strong>chat</strong> models for normal interaction</li><li><strong>embedding</strong> models for vector memory and search</li><li><strong>speech</strong> models for TTS</li></ul><p>Image generation is configured through compatible chat/image models rather than a separate image-provider system.</p><h3 id="operations-and-ui" tabindex="-1">Operations And UI <a class="header-anchor" href="#operations-and-ui" aria-label="Permalink to "Operations And UI""></a></h3><p>The Web UI is designed so you can manage the whole system without editing config files by hand every day. It includes:</p><ul><li>bot configuration tabs for general settings, access, channels, heartbeat, compaction, and more</li><li>provider and model management with OAuth flows where supported</li><li>session-side controls such as immediate compaction and status inspection</li><li>skill management with effective / shadowed / disabled visibility</li><li>slash-command driven control from channels</li></ul><h2 id="where-to-start" tabindex="-1">Where To Start <a class="header-anchor" href="#where-to-start" aria-label="Permalink to "Where To Start""></a></h2><ul><li><strong><a href="/installation/docker.html">Docker Installation</a></strong> — get the stack running</li><li><strong><a href="/getting-started/provider-and-model.html">Providers And Models</a></strong> — configure model access</li><li><strong><a href="/getting-started/bot.html">Bot Setup</a></strong> — create and configure a bot</li><li><strong><a href="/getting-started/sessions.html">Sessions</a></strong> — understand chat vs discuss behavior</li><li><strong><a href="/getting-started/channels.html">Channels</a></strong> — choose where bots are reachable</li><li><strong><a href="/getting-started/skills.html">Skills</a></strong> and <strong><a href="/getting-started/supermarket.html">Supermarket</a></strong> — extend what bots can do</li><li><strong><a href="/getting-started/slash-commands.html">Slash Commands</a></strong> — operate bots directly from chat</li></ul>',37)])])}const u=t(i,[["render",n]]);export{g as __pageData,u as default};
|