Name: Novel-Team
Author: woguaiwo

Imported from woguaiwo/Novel-Team (AGENTS.md). Install upstream with npx skills add woguaiwo/Novel-Team. Copyright stays with the author.

Novel-Team Agent 开发指南

项目结构约定

Python 代码遵循 PEP 8 规范。
所有提示词模板集中放在 novel_team/prompts/templates.py，便于统一调优。
智能体业务逻辑放在 novel_team/agents/，每个角色一个文件。
脚本入口放在 scripts/，避免与包内模块耦合。

添加新智能体的标准流程

编写系统提示词 → novel_team/prompts/templates.py
实现 Agent 类 → novel_team/agents/your_agent.py
- 继承 BaseAgent
- 实现 run(self, **kwargs) -> str
- 通过 self.ask(user_prompt) 调用 LLM
导出类 → novel_team/agents/__init__.py
编排进流水线 → novel_team/engine/team_orchestrator.py
更新 README 说明新角色职责

LLM 调用规范

禁止在 BaseAgent 之外直接调用 LLMEngine.generate。
所有 temperature / max_tokens 应通过配置文件读取，允许按角色差异化。
生成结果需做 .strip() 处理，去除首尾空白。

调试技巧

使用 scripts/test_chapter.py 快速验证新角色的提示词效果。
若显存不足，可在 config/novel_config.yaml 中将 model.torch_dtype 改为 float16 或尝试 load_in_8bit。
审查报告中的 【需修改】 关键字会触发自动重写逻辑，可在 TeamOrchestrator.write_chapter 中调整判定规则。

# after Install: the listing is in your workspace's routing pool - a plan picks it for its slot curl -s -X POST https://api.opensmartroute.ai/api/v1/route -H 'Authorization: Bearer $OSR_API_KEY' -H 'Content-Type: application/json' -d '{"text": "...", "plan": true}'

Install Novel-Team and it becomes one more candidate the router can pick - when it fits.

A listing is a routing target with a manifest: what it does, which domains it covers, what it costs and who publishes it. Once installed it sits beside your own models and tools, is scored like any other candidate for each request, and shows up in the trace when it wins. Ratings come from workspaces that installed it, one per account.

{ "ocm": "1", "id": "woguaiwo-novel-team-novel-team", "kind": "skill", "name": "Novel-Team", "description": "1. **编写系统提示词** → `novel_team/prompts/templates.py` 2. **实现 Agent 类** → `novel_team/agents/your_agent.py` - 继承 `BaseAgent` - 实现 `run(self, **kwargs) -> str` - 通过 `self.ask(user_prompt)` 调用 LLM 3. **导出类** → `novel_team/agents/__init__.py` 4. **编排进流水线** → `novel_team/engine/team_orchestrator.py` 5. **更新 README** 说明新角色职责", "publisher": "woguaiwo", "version": "1.0.0", "capabilities": { "domains": [ "creative" ], "tags": [ "skill-md", "github-agents" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "1. **编写系统提示词** → `novel_team/prompts/templates.py` 2. **实现 Agent 类** → `novel_team/agents/your_agent.py` - 继承 `BaseAgent` - 实现 `run(self, **kwargs) -> str` - 通过 `self.ask(user_prompt)` 调用 LLM 3. **导出类** → `novel_team/agents/__init__.py` 4. **编排进流水线** → `novel_team/engine/team_orchestrator.py` 5. **更新 README** 说明新角色职责" ], "primary": false, "metadata": { "source": { "provider": "github-agents", "repository": "https://github.com/woguaiwo/Novel-Team", "path": "AGENTS.md", "ref": "34d0bdf2d0516218ab98d31a2de45dee11987afe", "url": "https://github.com/woguaiwo/Novel-Team/blob/34d0bdf2d0516218ab98d31a2de45dee11987afe/AGENTS.md", "key": "woguaiwo/Novel-Team/AGENTS.md" } }, "instructions": "# Novel-Team Agent 开发指南\n\n## 项目结构约定\n\n- Python 代码遵循 PEP 8 规范。\n- 所有提示词模板集中放在 `novel_team/prompts/templates.py`，便于统一调优。\n- 智能体业务逻辑放在 `novel_team/agents/`，每个角色一个文件。\n- 脚本入口放在 `scripts/`，避免与包内模块耦合。\n\n## 添加新智能体的标准流程\n\n1. **编写系统提示词** → `novel_team/prompts/templates.py`\n2. **实现 Agent 类** → `novel_team/agents/your_agent.py`\n - 继承 `BaseAgent`\n - 实现 `run(self, **kwargs) -> str`\n - 通过 `self.ask(user_prompt)` 调用 LLM\n3. **导出类** → `novel_team/agents/__init__.py`\n4. **编排进流水线** → `novel_team/engine/team_orchestrator.py`\n5. **更新 README** 说明新角色职责\n\n## LLM 调用规范\n\n- 禁止在 `BaseAgent` 之外直接调用 `LLMEngine.generate`。\n- 所有", "cost": { "context_tokens": 225 } }

Novel-Team

About

Novel-Team Agent 开发指南

项目结构约定

添加新智能体的标准流程

LLM 调用规范

调试技巧

Use it

Install Novel-Team and it becomes one more candidate the router can pick - when it fits.

Manifest

Reviews