Skip to main content
Glama

topic_based_summary

Generate concise, topic-focused summaries from provided content using a targeted query. Ideal for extracting key insights and relevant information efficiently.

Instructions

主题汇总功能 - 基于给定资料和查询主题,返回最相关的内容总结(2k字符内) Args: content: 资料内容 query: 查询的主题或问题 Returns: 基于主题的相关内容总结

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
contentYes
queryYes

Implementation Reference

  • The primary handler for the 'topic_based_summary' tool, registered via @mcp.tool() decorator. It wraps the RAGProcessor's topic_summary method, handling context logging and errors.
    @mcp.tool() async def topic_based_summary(content: str, query: str, ctx: Context) -> str: """ 主题汇总功能 - 基于给定资料和查询主题,返回最相关的内容总结(2k字符内) Args: content: 资料内容 query: 查询的主题或问题 Returns: 基于主题的相关内容总结 """ try: ctx.info(f"开始主题汇总,查询: {query}") summary = await rag_processor.topic_summary(content, query) ctx.info("主题汇总完成") return summary except Exception as e: logger.error(f"主题汇总失败: {e}") return f"主题汇总失败: {str(e)}"
  • Core helper function in RAGProcessor class that performs the topic-based summarization by crafting a specialized prompt for the LLM and generating the response.
    async def topic_summary(self, content: str, query: str) -> str: """ 基于主题查询的内容总结 Args: content: 资料内容 query: 查询主题/问题 Returns: 相关内容的总结(2k字符内) """ try: # 构建RAG样式的提示词 prompt = f"""基于以下资料内容,针对查询主题进行总结分析: 查询主题: {query} 资料内容: {content} 请根据查询主题,从资料中提取最相关的信息,并总结为2000字以内的内容。要求: 1. 重点关注与查询主题相关的内容 2. 保持信息的准确性和逻辑性 3. 如果资料中没有相关信息,请明确说明 4. 提供具体的细节和要点""" response = self.summarizer.client.chat.completions.create( model=OPENAI_MODEL, messages=[ {"role": "system", "content": "你是一个专业的信息分析师,擅长从大量资料中提取特定主题的相关信息。"}, {"role": "user", "content": prompt} ], max_tokens=min(MAX_OUTPUT_TOKENS, 120000), # 限制在120k左右 temperature=0.1 ) result = response.choices[0].message.content.strip() return result except Exception as e: logger.error(f"主题汇总失败: {e}") return f"主题汇总失败: {str(e)}"

Other Tools

Related Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/yzfly/fullscope-mcp-server'

If you have feedback or need assistance with the MCP directory API, please join our Discord server