content-pipeline

Overview

Generates content through a structured pipeline: requirements discovery, research, parallel content generation, optional translation, quality validation, and final assembly.

Required References

Before starting, read references/writing/writing-plans.md for content structure, tone guidelines, and editorial workflow patterns.

Step 1: Discovery

1. Understand content requirements (type, audience, tone, length)
2. Identify brand voice and style guidelines
3. Determine output format (markdown, HTML, JSON)
4. Check for multilingual requirements

Step 2: Research

1. Analyze existing content for style and patterns
2. Gather domain knowledge relevant to the topic
3. Identify key messages and value propositions
4. Create content outline

Step 3: Generate (Parallel)

Generate content sections in parallel:

1. Each section gets its own generation pass
2. Follow the brand voice and style guidelines
3. Apply humanizer patterns from references/writing/humanizer.md to avoid AI-sounding text
4. Include SEO keywords where applicable

Step 4: Translate (Conditional)

Only if multilingual content is requested.

1. Translate content to target locales
2. Adapt idioms and cultural references (not word-for-word)
3. Preserve formatting and structure
4. Flag untranslatable terms

Step 5: Validate

1. Check content against requirements
2. Verify tone and voice consistency
3. Check for factual accuracy
4. Validate links and references
5. Run spelling and grammar checks

Step 6: Assemble

1. Combine all content sections into final output
2. Apply consistent formatting
3. Add metadata (author, date, tags)
4. Generate table of contents if applicable