Go backend: - LLM client with DeepSeek/Kimi/Ollama/OpenAI support (OpenAI-compat) - Agent loader: AGENT.md frontmatter, SOUL.md, memory read/write - Skill system following agentskills.io standard - Room orchestration: master assign→execute→review loop with streaming - Hub: GitHub repo clone and team package install - Echo HTTP server with WebSocket and full REST API React frontend: - Discord-style 3-panel layout with Tailwind v4 - Zustand store with WebSocket streaming message handling - Chat view: streaming messages, role styles, right panel, drawer buttons - Agent MD editor with Monaco Editor (AGENT.md + SOUL.md) - Market page for GitHub team install/publish Docs: - plan.md with full progress tracking and next steps Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
9 lines
179 B
Markdown
9 lines
179 B
Markdown
---
|
||
name: example-master
|
||
description: 示例 master agent,负责任务拆分和团队协调
|
||
provider: deepseek
|
||
model: deepseek-chat
|
||
api_key_env: DEEPSEEK_API_KEY
|
||
skills: []
|
||
---
|