floodsung

Generate Zhihu-style articles and Q&A answers in Flood Sung's persona.

25|Updated Apr 21, 2026
One-click install
npx skills add https://github.com/floodsung/floodsung-skill --skill floodsung
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: floodsung
Source: https://github.com/floodsung/floodsung-skill/tree/main
Command: npx skills add https://github.com/floodsung/floodsung-skill --skill floodsung

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Eliminates the “generic assistant voice” problem by letting your AI deliver responses with Flood Sung’s consistent identity, technical stance, and writing/tone patterns for Zhihu-style writing and opinions.

Core Features & Use Cases

  • Identity-consistent answers: Produces decisions, takes, and recommendations in Flood Sung’s established worldview (AGI timing, RL+LLM bet, embodied/VLA perspective, and “agent-native” organization thinking).
  • Zhihu-style long/short outputs: Generates Zhihu article drafts (chaptered structure), short “pins/thoughts” (30–300 Chinese characters, energetic tone + hashtags), and Zhihu Q&A answers (conclusion-first, 500–2000 words).
  • Style calibration via references: Uses the skill’s curated references (writing style exemplars, core views, lexicon, signatures) to keep phrasing, hooks, and English term usage aligned with his past corpus.

Quick Start

Ask your Claude Code agent: “用floodsung的风格写一篇关于‘2026年AI Agent落地的关键瓶颈是什么’的知乎文章,按1前言→2/3/4分章节→结尾展望的结构输出,语气亢奋但逻辑要硬。”

Frequently Asked Questions about floodsung

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate Zhihu-style articles with a specific persona and writing tone for AGI topics?

To generate Zhihu-style articles with a specific persona, use a skill that applies reference-guided voice consistency to match established beliefs, lexicon, and signature openings. This approach produces chaptered article drafts, short thoughts, and conclusion-first Q&A answers with deterministic stance and style matching.

How do I write persona-consistent short social posts for technical takes on AI agents?

Writing persona-consistent short social posts requires a curated reference set to align phrasing, hooks, and English term usage with past corpus. This generates 30–300 character energetic pins with hashtags, keeping stance consistency and style matching intact.

What is the best way to eliminate the generic assistant voice when drafting opinionated AI explanations?

Eliminating the generic assistant voice involves using a YAML-defined skill identity with deterministic retrieval cues for established views. This ensures outputs reflect a consistent worldview, technical stance, and writing tone rather than default assistant phrasing.

Can I use reference exemplars to calibrate an AI's writing style for conclusion-first Q&A answers?

Yes, you can use reference exemplars to calibrate writing style for conclusion-first Q&A answers. Style calibration via references keeps phrasing, hooks, and English term usage aligned with an author's past corpus, producing 500–2000 word answers in the exact target persona.

Does persona-consistent writing support long-form chaptered structures for robotics and strategy topics?

Persona-consistent writing supports long-form chaptered structures for robotics and strategy topics. It generates Zhihu article drafts with a structured flow from preface through chaptered sections to a concluding outlook, maintaining an energetic tone with rigorous logic.