ContentForge — Cross-Platform Content Distillation Skill

A Claude Code Skill that collects viral posts from X/Twitter, XHS (Xiaohongshu), and WeChat Official Accounts (GZH), distills top content by engagement scoring, synthesizes original posts via Claude, and learns from human edits to improve over time.

Pipeline

Collect → Distill → Synthesize → Review → Learn → Promote

Phase	What it does
Collect	Gather viral posts from X (bird CLI), XHS, GZH
Distill	Dedup + engagement scoring → top 30 sources
Synthesize	Claude generates platform-specific content
Review	Human edits, scores, and comments on drafts
Learn	Extract patterns from edits → update style rules
Promote	High-performing articles feed back as examples

Install

# Clone into your Claude Code skills directory
git clone https://github.com/SamCuipogobongo/content-forge-skill.git ~/.claude/skills/content-forge

Or download the .skill package from Releases.

Prerequisites

• Node.js >= 18
• bird CLI — for X/Twitter search (only if using X platform)
• Python 3 — for XHS/GZH scrapers (only if using those platforms)
• Claude CLI — claude command in PATH

Usage

# Generate content for a keyword (X only)
node scripts/content-forge-pipeline.js --keyword "AI agents" --platforms x

# All 3 platforms
node scripts/content-forge-pipeline.js --keyword "AI agents" --platforms x,xhs,gzh

# Learn from yesterday's human edits
node scripts/content-forge-pipeline.js --learn

# Check recent runs
node scripts/content-forge-pipeline.js --status

Configuration

Env Variable	Default	Description
BIRD_BIN	bird	bird CLI path
PYTHON_BIN	python3	Python interpreter
CLAUDE_BIN	claude	Claude CLI binary
MIN_FOLLOWERS	1000	Min followers for X source filtering
CONTENT_FORGE_DATA_DIR	./content-forge-data	Pipeline data directory
SOCIAL_MEDIA_DIR	./Social Media	Platform article copies

Self-Learning Loop

1. Run the pipeline → generates .md files with draft content
2. Human edits the drafts inline (original vs edited sections)
3. Run --learn → Claude analyzes edit patterns
4. Style rules auto-update in references/style-rules-{platform}.md
5. Next run uses improved rules → better content

License

MIT