Content Creation and Repurposing

The Content Operations Problem

Most people who create content are leaving enormous value on the table. They record a podcast episode that has 45 minutes of insight and extract it into one audio file. They write a detailed client proposal that contains paragraphs that would make excellent LinkedIn posts. They run a customer call with quotable moments that never become marketing material.

The bottleneck isn't ideas — it's bandwidth. Repurposing content is mechanical, repetitive work: transcribing, summarizing, reformatting, adapting for different platforms, maintaining consistent voice across formats. It's exactly the kind of work Co-Work was designed to eliminate.

Garbage In, Treasure Out

Jenny's framing inverts the traditional computing axiom. In standard software, bad input produces bad output — GIGO. In Co-Work content work, the messy inputs that would defeat traditional tools are exactly where Co-Work excels. Raw transcripts with filler words. Meeting recordings where three people talk over each other. Rough notes that are half-formed ideas. PDF reports with dense data tables.

Co-Work doesn't need clean input. It can extract signal from noise. That means your starting material doesn't need to be polished — it needs to be present. The synthesis is Co-Work's job.

The Repurposing Pipeline Architecture

The fundamental pattern: one input source, processed by Co-Work, producing multiple simultaneous outputs tailored for different contexts.

Input sources:

• YouTube URL — Co-Work extracts the transcript via YouTube connector or Chrome
• Meeting recording — upload or reference via Drive; Co-Work transcribes and analyzes
• PDF or document — uploaded or referenced from Drive; Co-Work reads and extracts
• Web article URL — Co-Work reads via Chrome or web connector
• Raw notes — pasted directly into the conversation or saved in the content folder

Output formats (all generated from the same input in one pass):

• LinkedIn post — professional, voice-matched, 150–300 words
• Twitter/X thread — headline hook + 5–7 supporting points
• Email newsletter paragraph — summary for existing subscriber context
• Blog post outline — H2 structure with key points per section
• Slide deck outline — 5–8 slides with titles and bullet points
• Video script intro — 60-second hook version of the same content

Brooke Wright built and documented this end-to-end repurposing workflow. Paul extended it into a full content OS. The key insight they share: Co-Work generates all formats simultaneously, not sequentially. You're not running one prompt per format — you're running one repurposing skill that outputs everything at once.

YouTube Transcript Extraction with Timestamps

For video content, Jack Roberts recommends extracting transcripts with two levels of timestamps: chapter-level (by major topic) and minute-level (for precise quotes). The prompt structure:

Extract the transcript from [YouTube URL].
Format: Chapter headings with timestamps, then
line-by-line transcript with minute markers.
Identify the top 5 most quotable moments with exact text.

The output gives you: timestamped show notes for the video description, specific quotes for social posts, chapter titles for long-form content, and a searchable reference document for future repurposing runs.

Brand Voice Preservation

Automated content generation has one persistent failure mode: it sounds like it was generated by a machine. The technical execution is correct, but the voice is generic. This destroys the value for anyone who has spent years building a recognizable communication style.

Brooke Wright's solution is the voice reference document. The process:

1. Identify your 3–5 best existing content pieces — the ones that best represent your authentic voice
2. Create a file called voice-reference.md in your Content folder
3. Add those examples in full, with a note about what makes each one work
4. When creating your Content Repurposer skill, feed these examples to the Skill Creator
5. Add to the skill instructions: "Match the voice, sentence length, and energy of the examples in voice-reference.md"

The Monday Morning Insights Ritual

Jenny describes her Monday morning ritual as a three-prompt sequence that turns a week's worth of inputs into actionable priorities plus a shareable presentation:

1. Multi-source synthesis: Pull in data from multiple sources (user research, analytics, customer feedback, market signals) and synthesize into key themes
2. Feature priorities: From the themes, extract what the team should focus on this week and why
3. Presentation: "These are my insights. Create a presentation with 8 slides." Export to Google Slides via connector

This same pattern applies far beyond Monday mornings. Quarterly business reviews. Client check-ins. Research summaries. Any workflow where the value is in synthesizing multiple sources into a coherent narrative and then formatting it for presentation.

Presentation Creation

When Co-Work generates a slide deck outline, the natural next step is exporting to a real presentation. With the Google Slides connector active, the prompt is straightforward:

Take these insights and create a presentation
with 8 slides in Google Slides.
Title slide, 6 content slides with bullet points,
closing next-steps slide.

Four sources — Jenny, Paul, Jack Roberts, and Brooke Wright — all describe presentation generation as one of the highest-ROI Co-Work workflows. The manual version takes an hour or more; the automated version takes a few minutes of review and editing.

Content Operations OS

Paul's content OS is the full stack version of what this module introduces. It combines everything:

• A dedicated ~/CoWork/Content/ folder with a voice-reference document and content-specific claude.md
• A Content Repurposer skill that takes any input format and produces all outputs simultaneously
• A Voice Checker skill that reviews outputs against the voice reference before finalizing
• A Content Calendar skill that suggests what to publish and when based on pipeline state
• Scheduled weekly runs that process accumulated inputs and surface content ideas