content-generation-example
v0.1.0
Published
Multi-format content generation pipeline — research, draft, 4 parallel formats, optional translate
Readme
Multi-Format Content Generation
Generates 4 content formats simultaneously from a single topic using flomatai's parallel step. Research a topic once, produce a blog post, Twitter thread, LinkedIn post, and TL;DR newsletter snippet in parallel, with optional translation.
Pipeline Graph
research-topic
↓
parallel (all 4 formats at once):
├── format-blog → blog post (600+ words, markdown)
├── format-twitter → thread (5-8 tweets, ≤280 chars each)
├── format-linkedin → post (1300-1800 chars, professional)
└── format-tldr → snippet (headline + 3 bullets + takeaway)
↓
[optional] translate-content (--translate-to <language>)Setup
cp .env.example .env
# Fill in ANTHROPIC_API_KEY (or OPENAI_API_KEY)
pnpm install
pnpm buildUsage
# Generate all 4 formats for a topic
node dist/src/run.js --topic "The future of AI automation"
# With audience and tone guidance
node dist/src/run.js \
--topic "TypeScript generics explained" \
--audience "junior developers" \
--tone "educational and friendly"
# With translation
node dist/src/run.js \
--topic "Remote work best practices" \
--translate-to "Spanish"
# Custom output directory
node dist/src/run.js --topic "..." --output ./my-contentOutput Files
All files are saved to ./output/ (or --output path):
| File | Description |
|------|-------------|
| <slug>-blog.md | Long-form blog post in Markdown |
| <slug>-twitter.md | Twitter thread (numbered tweets) |
| <slug>-linkedin.md | LinkedIn post ready to publish |
| <slug>-tldr.md | Newsletter snippet |
| <slug>-blog-<lang>.md | Translated blog (if --translate-to set) |
Environment Variables
| Variable | Description |
|----------|-------------|
| ANTHROPIC_API_KEY | Anthropic API key |
| OPENAI_API_KEY | OpenAI API key (alternative) |
| OPENCODE_BASE_URL | Use local OpenCode proxy |
| OUTPUT_DIR | Output directory (default: ./output) |
