feat: social layer — messaging, Hire/Visit relationships, contacts, notifications

.env.example

@@ -16,19 +16,3 @@ TAVILY_API_KEY=your-tavily-api-key
 
 # LangSmith tracing (optional)
 LANGSMITH_API_KEY=your-langsmith-api-key
-
-# Supabase (required when LEON_STORAGE_STRATEGY=supabase)
-LEON_STORAGE_STRATEGY=supabase
-SUPABASE_PUBLIC_URL=https://supabase.mycel.nextmind.space
-
-# SUPABASE_INTERNAL_URL: direct server-side URL (bypasses public proxy).
-# Production (same-host): SUPABASE_INTERNAL_URL=http://<internal-host>:8000
-# Local dev (SSH tunnel):  SUPABASE_INTERNAL_URL=http://localhost:18000
-SUPABASE_INTERNAL_URL=http://localhost:18000
-
-SUPABASE_ANON_KEY=your-anon-key
-LEON_SUPABASE_SERVICE_ROLE_KEY=your-service-role-key
-SUPABASE_JWT_SECRET=your-jwt-secret
-
-# DB schema: staging for local dev and staging envs; omit for production (defaults to public)
-LEON_DB_SCHEMA=staging

.github/workflows/publish.yml

@@ -0,0 +1,39 @@
+name: Publish to PyPI
+
+on:
+  push:
+    tags:
+      - 'v*'
+  workflow_dispatch:
+
+jobs:
+  build-and-publish:
+    runs-on: ubuntu-latest
+    permissions:
+      id-token: write
+      contents: read
+
+    steps:
+    - uses: actions/checkout@v4
+
+    - name: Set up Python
+      uses: actions/setup-python@v5
+      with:
+        python-version: '3.12'
+
+    - name: Install build dependencies
+      run: |
+        python -m pip install --upgrade pip
+        pip install build twine
+
+    - name: Build package
+      run: python -m build
+
+    - name: Check package
+      run: twine check dist/*
+
+    - name: Publish to PyPI
+      env:
+        TWINE_USERNAME: __token__
+        TWINE_PASSWORD: ${{ secrets.PYPI_TOKEN }}
+      run: twine upload dist/*

.gitignore

@@ -46,8 +46,6 @@ frontend/app/.env.development
 .claude/.stfolder/
 .claude/.vscode/
 .claude/settings.local.json
-.claude/mcp.json
-.mcp.json
 teams
 
 # User-level Leon config and skills
@@ -94,7 +92,6 @@ worktrees/
 /*.png
 /*.yml
 /*.yaml
-!docker-compose.yml
 /dogfood-output/
 /current-chat.yaml
 /.claude/skills/
@@ -106,6 +103,5 @@ frontend/.vite/
 .playwright-cli/
 ops
 
-# Auto-generated
+.lark-events/
 .playwright-mcp/
-/supabase/

README.md

@@ -4,7 +4,7 @@
 
 <img src="./assets/banner.png" alt="Mycel Banner" width="600">
 
-**Link: connecting people, agents, and teams for the next era of human-AI collaboration**
+**Production-ready agent runtime for building, running, and governing collaborative AI teams**
 
 🇬🇧 English | [🇨🇳 中文](README.zh.md)
 
@@ -15,16 +15,16 @@
 
 ---
 
-Mycel gives your agents a **body** (portable identity & sandbox), **mind** (shareable templates), **memory** (persistent context), and **social life** (a native messaging layer where humans and agents coexist as equals). It's the platform layer for human-AI teams that actually work together.
+Mycel is an enterprise-grade agent runtime that treats AI agents as long-running co-workers. Built on a middleware-first architecture, it provides the infrastructure layer missing from existing agent frameworks: sandbox isolation, multi-agent communication, and production governance.
 
 ## Why Mycel?
 
-Existing frameworks help you *build* agents. Mycel helps agents *live* — move between tasks, accumulate knowledge, message teammates, and collaborate in workflows that feel as natural as a group chat.
+Existing agent frameworks focus on *building* agents. Mycel focuses on *running* them in production:
 
-- **Body** — Agents get a portable identity with sandbox isolation. Deploy anywhere (Local, Docker, E2B, Daytona, AgentBay), migrate seamlessly, and let your agents work for you — or for others.
-- **Mind** — A template marketplace for agent personas and skills. Share your agent's configuration, subscribe to community templates, or let a well-designed agent earn its keep.
-- **Memory** — Persistent, structured memory that travels with the agent across sessions and contexts.
-- **Social** — All members of the platform — human or AI — exist as first-class entities. Chat naturally, share files, forward conversation threads to agents: the social graph is the collaboration layer.
+- **Middleware Pipeline**: Unified tool injection, validation, security, and observability
+- **Sandbox Isolation**: Run agents in Docker/E2B/cloud with automatic state management
+- **Multi-Agent Communication**: Agents discover, message, and collaborate with each other — and with humans
+- **Production Governance**: Built-in security controls, audit logging, and cost tracking
 
 ## Quick Start
 
@@ -59,7 +59,7 @@ uv sync --extra e2b         # E2B
 uv sync --extra daytona     # Daytona
 ```
 
-Docker sandbox works out of the box (just needs Docker installed). See [Sandbox docs](docs/en/sandbox.mdx) for provider setup.
+Docker sandbox works out of the box (just needs Docker installed). See [Sandbox docs](docs/en/sandbox.md) for provider setup.
 
 ### 3. Start the services
 
@@ -170,11 +170,12 @@ Agents can be extended with external tools and specialized expertise:
 
 ## Documentation
 
-- [Configuration](docs/en/configuration.mdx) — Config files, virtual models, tool settings
-- [Multi-Agent Chat](docs/en/multi-agent-chat.mdx) — Entity-Chat system, agent communication
-- [Sandbox](docs/en/sandbox.mdx) — Providers, lifecycle, session management
-- [Deployment](docs/en/deployment.mdx) — Production deployment guide
-- [Concepts](docs/en/concepts.mdx) — Core abstractions (Thread, Member, Task, Resource)
+- [CLI Reference](docs/en/cli.md) — Terminal interface, commands, LLM provider setup
+- [Configuration](docs/en/configuration.md) — Config files, virtual models, tool settings
+- [Multi-Agent Chat](docs/en/multi-agent-chat.md) — Entity-Chat system, agent communication
+- [Sandbox](docs/en/sandbox.md) — Providers, lifecycle, session management
+- [Deployment](docs/en/deployment.md) — Production deployment guide
+- [Concepts](docs/en/product-primitives.md) — Core abstractions (Thread, Member, Task, Resource)
 
 ## Contact Us
 

README.zh.md

@@ -4,7 +4,7 @@
 
 <img src="./assets/banner.png" alt="Mycel Banner" width="600">
 
-**Link：连接人与 Agent，构建下一代人机协同**
+**企业级 Agent 运行时，构建、运行和治理协作 AI 团队**
 
 [🇬🇧 English](README.md) | 🇨🇳 中文
 
@@ -15,16 +15,16 @@
 
 ---
 
-Mycel 让你的 Agent 拥有**身体**（可迁移的身份与沙箱）、**思想**（可共享的模板市场）、**记忆**（跨会话的持久上下文）和**社交**（人与 Agent 平等共存的原生消息层）。这是真正意义上的人机协同平台。
+Mycel 是企业级 Agent 运行时，将 AI Agent 视为长期运行的协作伙伴。基于中间件优先架构，提供现有 Agent 框架缺失的基础设施层：沙箱隔离、多 Agent 通讯和生产治理。
 
 ## 为什么选择 Mycel？
 
-现有框架帮你*构建* Agent，Mycel 让 Agent 真正*活着*——在任务间自由迁移、积累知识、给队友发消息，用像群聊一样自然的方式协作。
+现有 Agent 框架专注于*构建* Agent，Mycel 专注于在生产环境*运行*它们：
 
-- **身体** — Agent 拥有可迁移的身份和沙箱隔离。支持 Local / Docker / E2B / Daytona / AgentBay，随时迁移，让你的 Agent 为你工作，也能为别人打工。
-- **思想** — Agent 模板市场：分享你的 Agent 配置，订阅社区模板，让设计精良的 Agent 产生真实价值。
-- **记忆** — 持久结构化记忆，跟随 Agent 跨会话、跨上下文流转。
-- **社交** — 平台上所有成员——无论是人还是 AI——都是一等公民实体。像微信一样自然地聊天、发文件、把聊天记录分享给 Agent：社交图谱就是协作层。
+- **中间件管线**：统一的工具注入、校验、安全和可观测性
+- **沙箱隔离**：在 Docker/E2B/云端运行 Agent，自动状态管理
+- **多 Agent 通讯**：Agent 之间互相发现、发送消息、自主协作——人类也参与其中
+- **生产治理**：内置安全控制、审计日志和成本追踪
 
 ## 快速开始
 
@@ -59,7 +59,7 @@ uv sync --extra e2b         # E2B
 uv sync --extra daytona     # Daytona
 ```
 
-Docker 沙箱开箱即用（只需安装 Docker）。详见[沙箱文档](docs/zh/sandbox.mdx)。
+Docker 沙箱开箱即用（只需安装 Docker）。详见[沙箱文档](docs/zh/sandbox.md)。
 
 ### 3. 启动服务
 
@@ -170,11 +170,12 @@ Agent 可通过外部工具和专业技能进行扩展：
 
 ## 文档
 
-- [配置指南](docs/zh/configuration.mdx) — 配置文件、虚拟模型、工具设置
-- [多 Agent 通讯](docs/zh/multi-agent-chat.mdx) — Entity-Chat 系统、Agent 间通讯
-- [沙箱](docs/zh/sandbox.mdx) — 提供商、生命周期、会话管理
-- [部署](docs/zh/deployment.mdx) — 生产部署指南
-- [核心概念](docs/zh/concepts.mdx) — 核心抽象（Thread、Member、Task、Resource）
+- [CLI 参考](docs/zh/cli.md) — 终端界面、命令、LLM 提供商配置
+- [配置指南](docs/zh/configuration.md) — 配置文件、虚拟模型、工具设置
+- [多 Agent 通讯](docs/zh/multi-agent-chat.md) — Entity-Chat 系统、Agent 间通讯
+- [沙箱](docs/zh/sandbox.md) — 提供商、生命周期、会话管理
+- [部署](docs/zh/deployment.md) — 生产部署指南
+- [核心概念](docs/zh/product-primitives.md) — 核心抽象（Thread、Member、Task、Resource）
 
 ## 联系我们
 

backend/taskboard/service.py

@@ -217,7 +217,7 @@ def _get_thread_id(self) -> str:
     # Handlers (async — ToolRunner awaits coroutines)
     # ------------------------------------------------------------------
 
-    async def _list_tasks(self, Status: str = "", Priority: str = "") -> str:
+    async def _list_tasks(self, Status: str = "", Priority: str = "") -> str:  # noqa: N803
         try:
             tasks = await asyncio.to_thread(task_service.list_tasks)
         except Exception as e:
@@ -231,7 +231,7 @@ async def _list_tasks(self, Status: str = "", Priority: str = "") -> str:
 
         return json.dumps({"tasks": tasks, "total": len(tasks)}, ensure_ascii=False)
 
-    async def _claim_task(self, TaskId: str) -> str:
+    async def _claim_task(self, TaskId: str) -> str:  # noqa: N803
         thread_id = self._get_thread_id()
         now_ms = int(time.time() * 1000)
         try:
@@ -249,7 +249,7 @@ async def _claim_task(self, TaskId: str) -> str:
             return json.dumps({"error": f"Task not found: {TaskId}"})
         return json.dumps({"task": updated}, ensure_ascii=False)
 
-    async def _update_progress(self, TaskId: str, Progress: int, Note: str = "") -> str:
+    async def _update_progress(self, TaskId: str, Progress: int, Note: str = "") -> str:  # noqa: N803
         update_kwargs: dict[str, Any] = {"progress": Progress}
 
         if Note:
@@ -272,7 +272,7 @@ async def _update_progress(self, TaskId: str, Progress: int, Note: str = "") ->
             return json.dumps({"error": f"Task not found: {TaskId}"})
         return json.dumps({"task": updated}, ensure_ascii=False)
 
-    async def _complete_task(self, TaskId: str, Result: str) -> str:
+    async def _complete_task(self, TaskId: str, Result: str) -> str:  # noqa: N803
         now_ms = int(time.time() * 1000)
         try:
             updated = await asyncio.to_thread(
@@ -290,7 +290,7 @@ async def _complete_task(self, TaskId: str, Result: str) -> str:
             return json.dumps({"error": f"Task not found: {TaskId}"})
         return json.dumps({"task": updated}, ensure_ascii=False)
 
-    async def _fail_task(self, TaskId: str, Reason: str) -> str:
+    async def _fail_task(self, TaskId: str, Reason: str) -> str:  # noqa: N803
         now_ms = int(time.time() * 1000)
         try:
             updated = await asyncio.to_thread(
@@ -307,7 +307,7 @@ async def _fail_task(self, TaskId: str, Reason: str) -> str:
             return json.dumps({"error": f"Task not found: {TaskId}"})
         return json.dumps({"task": updated}, ensure_ascii=False)
 
-    async def _create_task(self, Title: str, Description: str = "", Priority: str = "medium") -> str:
+    async def _create_task(self, Title: str, Description: str = "", Priority: str = "medium") -> str:  # noqa: N803
         try:
             task = await asyncio.to_thread(
                 task_service.create_task,