$ loading_
57690 Skills & MCP tools — copy a prompt, zero pre-install, into Claude Code / Codex / Cursor.
🎓 Skill = teach the AI how to do a task, copy & use; 🔌 MCP = connect the AI to an external tool/data source, usually needs a key.
Demonstrate MCP tools, resources, and prompts for comprehensive protocol capability testing.
Helps AI break down complex problems step by step and reflect on its reasoning.
Store and recall cross-conversation details with persistent knowledge-graph-based memory.
Let AI read, search, and manage local Git repositories and commit history.
Fetch web pages and convert them to Markdown for AI reading and analysis.
Securely read and write local files, manage directories, and quickly search needed content.
Retrieve the latest library documentation with version-matched code examples for reference.
Automate browser interactions, capture screenshots, and run end-to-end testing workflows.
Read Figma designs and quickly help turn interface structures into code.
Crawl websites with JS rendering and extract structured data for content collection and analysis.
Connect to and manage Supabase projects, perform table operations, and analyze queries.
Connect to PostgreSQL to inspect schemas and read or write business data.
Connect to Stripe APIs to handle payments, orders, and finance-related operations.
Send and receive Slack messages, manage channels, and handle workspace collaboration tasks.
Search the web and local results privately to quickly find information and resources.
Automate web interactions, scraping, and testing tasks using a headless browser.
Connect to SQLite databases, inspect table schemas, and safely read or write data.
Build quick throwaway prototypes to validate options and deliver a verdict.
Fetch GitHub issues, create fixes, open PRs, and handle reviews.
Create, search, and manage Bear notes on macOS via grizzly CLI.
Summarize CodexBar local cost logs by model for Codex or Claude.
Maintain an OpenClaw memory wiki with deterministic, source-backed page updates.
Debug Python issues with pdb, breakpoints, post-mortem analysis, and debugpy.
Triage security advisories, drafts, and GHSA reports with trusted evidence.
从访谈内容留存、转写回看，到主题归纳与产品建议，一套串起用户研究整理流程。
从流量采集、营销归因到数仓查询与看板输出，一套串起自动取数和分析闭环。
从读取设计稿、提取上下文到映射组件库与同步评审，这套组合很适合设计到前端落地的协作链路。
In-depth guides on Skill / MCP setup and security — beyond the directory listings
普通聊天 AI 很难真正读懂大型仓库，因为它缺少结构、上下文和可验证的检索路径。本文从 0 到 1 搭一个只面向本地代码库的 Cursor 助手：能查结构、追调用链、看 diff、给重构建议，还能尽量压住幻觉。
Agent-hop 是一个基于 Rust 的 TUI，支持在进行中的 Claude Code 聊天里切换到其他编程代理，尽量不丢失上下文。项目作者提到，它最早源自一个简历与搜索工具，后来因为使用过程中的阻碍，单独做成了现在的界面。
Harden.run 表示，针对能够生成任意代码的 Coding Agent，单靠传统红队测试并不够。他们通过后训练网络安全小模型、调整推理方式，并结合 inline reference monitoring 等程序分析手段，在 LinuxArena 和 SleightBench 上取得了超过 GPT5.5-xhigh 的成绩。