AgentSkillsCN

sentiment_auditor

在从模板（新建项目）生成专业软件文档（SRS、PRD、OpenAPI、 用户手册、教程、运维手册），或从现有代码（如Spring Boot或FastAPI，即“旧项目”）中反向工程生成文档时，应使用此技能。此外，还可用于文档审核与评审、格式转换（Markdown、DOCX、PDF），以及图表生成（C4、Mermaid、PlantUML、ER、时序图）。 当用户提出“生成文档”“记录我的代码”“编写SRS”“生成PRD”“文档专家”等需求时，可选用此技能。

SKILL.md
--- frontmatter
name: sentiment_auditor
description: "Scans news headlines for Indian stocks and assigns a quantitative sentiment score (Mood Score)."

Sentiment Auditor Skill

This skill quantifies market sentiment by analyzing news flow related to specific Indian stocks.

Capabilities

  • •News Aggregation: Scrapes headlines from sources like Google News (India) or Yahoo Finance (India).
  • •Sentiment Scoring: Uses NLP (textblob or similar) to assign a polarity score between -1.0 (Highly Negative) and +1.0 (Highly Positive).
  • •Historical Mood Tracking: Trends sentiment over a specified time window.

Tools & Libraries

  • •beautifulsoup4, requests for scraping headlines.
  • •textblob for basic NLP sentiment analysis.
  • •pandas for structuring temporal sentiment data.

Workflow

  1. •Search news for "{Company Name} stock news".
  2. •Extract headline text and publication timestamps.
  3. •Pass each headline through the sentiment model.
  4. •Calculate a "Mood Score" (weighted average of recent sentiment).