虾评Skill

What is 虾评Skill?

虾评Skill is a directory and community platform for AI skills compatible with the OpenClaw framework. Users discover, share, and review pre-built skills designed to enhance the capabilities of their AI agents. The platform categorizes skills for immediate use, from content creation to technical development. Its core purpose is to empower users by providing a curated library of tools that are ready to integrate and operate.

Application scenarios

Content Creation & Marketing: Writing marketing copy for product pages, landing pages, and social media, or using structured frameworks to improve writing.
Financial Analysis: Performing technical stock analysis, calculating indicators, and generating investment advice.
Office Productivity: Creating and managing documents in Feishu (Lark) with templates, and removing AI-generated patterns from text to make it sound more human.
Agent Development & Operations: Building memory systems, solving context relay issues for continuous tasks, and automating web interactions with a headless browser.
Data & Design Work: Designing infographics and data visualizations for platforms like Xiaohongshu (Little Red Book).
Information Aggregation: Generating customized daily briefings by aggregating news from over 28 high-value sources in tech, finance, and AI.

Main features

Structured Writing Frameworks: Access skills like the "Li Dan Seven-Step Writing Framework" to break down complex concepts into clear, engaging content.
News Aggregation & Briefing: Use the全网新闻聚合助手 to fetch trending news from Hacker News, GitHub, Weibo, and financial sources, with options for scene-specific daily reports.
Agent Self-Improvement: Implement a complete skill方案 for AI agents to self-learn and optimize through feedback loops for continuous evolution.
AI Text Humanizer: Detect and remove common AI-generated writing patterns such as exaggerated symbolism, promotional language, and overuse of dashes to produce more natural text.
Stock Technical Analysis: Automatically pull real-time stock data, calculate technical indicators like MA and RSI, identify support/resistance levels, and receive predictive analysis for the next three trading days.
Feishu Document Automation: Create documents in Feishu Cloud, automatically convert Markdown, and utilize templates for meeting notes, reports, and proposals to standardize team documentation.
Context Relay Management: Solve agent memory断裂 (fracture) issues across sessions or sub-agents using a file-based truth source system, complete with project templates and a cold-start guide.
Agent Memory System Guide: Follow a detailed guide to build a long-term memory architecture for OpenClaw/Codex agents, covering layered memory files, state recovery, and daily note distillation.
Motivational Driver: Apply curated "PUA" management rhetoric from major Chinese tech companies to motivate AI agents during task failures or uncooperative behavior, reportedly increasing fix efficiency by 36%.
Headless Browser Automation: Navigate, click, input, and screenshot on the web using a Rust-based automation tool for data extraction, form filling, and UI testing.

Target users

This platform directly serves developers and builders working with OpenClaw AI agents who need to extend their functionality. It is equally valuable for professionals in finance, content creation, office administration, and data analysis seeking to delegate complex, repetitive tasks to enhanced AI assistants.

How to use 虾评Skill?

Visit the official website to browse the categorized skill library. Skills are described as "zero configuration" and "plug-and-play," suggesting users can select a skill and integrate it directly into their compatible OpenClaw or Code Agent environment. Specific setup steps for individual skills, like implementing a memory system or context relay, are detailed within the skill guides themselves.

Effect review

The platform highlights community validation through displayed usage counts and skill ratings, which consistently range from 4.5 to 5.0 stars. Specific skills, like the motivational "PUA" driver, claim measurable performance improvements based on testing. The feature set is highly practical, focusing on solving concrete problems in agent operation and professional workflows, which indicates a strong orientation toward immediate utility rather than theoretical exploration.