huashu-material-search — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited huashu-material-search (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Quickly search Huashu's real experiences, opinions, and case studies to add authenticity and credibility to content creation.
This skill is automatically loaded when:
Suitable for:
Steps:
pattern: "keyword1|keyword2|keyword3"
path: /Users/alchain/Documents/writing/writing-reference/personal-material-library/all-jike-posts.csv
output_mode: content
-n: true
-i: true # Case-insensitiveFrom the search results, find:
Search examples:
# Writing about Amap street rankings
pattern: "Amap|street rankings"
# Writing about AI coding tool comparisons
pattern: "Cursor|Claude Code|Copilot|GitHub"
# Writing about product growth strategy
pattern: "growth|conversion|users|retention"
# Writing about technical learning experience
pattern: "learning|mastering|understanding|lessons learned"
# Writing about social media operations
pattern: "WeChat|video|Xiaohongshu|views|readership"Advantages:
Suitable for:
Steps:
/writing-reference/personal-material-library/README.mdOpen _jike-topic-index.md and find the relevant theme:
View the extracted file (e.g., AI-coding-tool-insights.md)
Follow _jike-usage-guidelines.md:
Available curated themes:
AI-coding-tool-insights.mdproduct-development-experience.mdsocial-media-operations-practice.mdtechnical-learning-methods.mdRequirement: Writing "Claude Code In-Depth Review" — need an opening
Search:
pattern: "Claude Code"Material found:
"Claude Code's biggest advantage is contextual understanding. When modifying a feature, it knows which files need to be changed."
Rewrite example:
Two weeks with Claude Code. The most noticeable change: previously, modifying a feature meant manually checking 5–6 files. Now Claude Code finds them all in one go. That ability is critical — especially when inheriting legacy projects.
Today, let's talk about whether Claude Code is worth it.Requirement: Writing "AI Coding Tool Selection Guide" — need a real comparison
Search:
pattern: "Cursor|Claude Code|comparison"Material found:
"Cursor suits small projects; Claude Code suits large projects. Price is the dealbreaker."
Rewrite example:
I used Cursor and Claude Code simultaneously for 3 months. My conclusion:
- Cursor: Good for personal projects and small teams. £15/month — good value for money.
- Claude Code: Better for large projects and complex refactors. £75+/month — more suited to companies.
If you're an individual developer, start with Cursor. If your company covers the expense, go for Claude Code.Requirement: Writing "Getting Started with HarmonyOS Development" — need a real project case
Search:
pattern: "HarmonyOS"Material found:
"HarmonyOS development's biggest pitfall is incomplete documentation. The official examples don't run."
Rewrite example:
I previously developed a HarmonyOS app using Cursor and hit quite a few snags. The biggest problem was incomplete documentation — about 30% of the official examples wouldn't run, and the error messages were vague.
Eventually I discovered the fastest approach is to read the source code directly. HarmonyOS component source code is fairly well-written — more reliable than the documentation.Requirement: Writing "The Value of Programmers in the AI Era" — need an elevated closing
Search:
pattern: "AI|replacement|programmer|value"Material found:
"AI can write for me, but it can't think for me. Real value lies in judgement."
Rewrite example:
AI can write code for me, but it can't think through why I need to write this code.
Technology will keep getting stronger, but judgement will always be human — choosing the tech stack, deciding what product to build, defining who to serve, and figuring out how to grow. That's something AI can't learn.
A programmer's value has never been the speed of typing code. It's the clarity of knowing what code needs to be written.Original material (Jike post):
"Amap's forced pop-up for the National Holiday street ranking feature. Absolutely disgusting. Product manager must have lost the plot."
❌ Incorrect rewriting (over-polished):
"Amap Maps still has room for improvement in user experience, particularly in the design of pop-up features, which could be made more user-friendly."
✅ Correct rewriting (preserves attitude, expands detail):
"Open Amap during the National Holiday and a pop-up forces you into the street ranking activity — you can't dismiss it without tapping 'Join'. I just want to navigate. Why am I being forced to join an activity?
The design logic here baffles me. What is the user's core need when opening a navigation app? To start navigating quickly — not to join a street ranking. A forced pop-up only creates resentment, and it won't help activity participation rates either. Will someone who was forced to tap 'Join' actually go out and walk the streets?
The core of product design is respecting user intent — not pushing KPIs on them."
A:
A:
A:
A:
# Search keywords (supports regex)
grep -i "Cursor\|Claude Code" /Users/alchain/Documents/writing/writing-reference/personal-material-library/all-jike-posts.csv
# Search with line numbers
grep -in "growth\|conversion" /Users/alchain/Documents/writing/writing-reference/personal-material-library/all-jike-posts.csv
# Search with context (3 lines before and after)
grep -i -C 3 "HarmonyOS" /Users/alchain/Documents/writing/writing-reference/personal-material-library/all-jike-posts.csvGrep tool parameters:
pattern: "keyword1|keyword2"
path: /Users/alchain/Documents/writing/writing-reference/personal-material-library/all-jike-posts.csv
output_mode: content
-n: true
-i: true
-C: 2 # Show context lines/writing-reference/personal-material-library/README.md — Master index (start here)/writing-reference/personal-material-library/all-jike-posts.csv — Raw data (1,800+ entries)/writing-reference/personal-material-library/_jike-topic-index.md — Topic categories/writing-reference/personal-material-library/_jike-usage-guidelines.md — Rewriting guidelines and examplesLast Updated: 2025-11-07 Applicable Projects: WeChat article writing, video scripting Maintained by: Huashu
By Huashu | AI Native Coder · Independent Developer WeChat Official Account "Huashu" | 300k+ followers | AI Tools & Productivity Flagship products: Kitty Catchlight (AppStore Paid Rankings Top 1) · Play DeepSeek with One Book
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.