Playbook — Blog Repurposing Engine (“Naval Ravikant bot”)

Summary

A built-and-running agent system that takes finished long-form we already own (a blog series, articles) and mines it into atomic, postable visual content. Agents read the article → write a brief → make a poster → insert that poster back into the article → post the poster with an excerpt. Mentally it’s a “Naval Ravikant bot”: turn a deep body of writing into an endless stream of quotable, on-brand micro-content. A Practical Market / Distribution tactic that runs daily.

What it is

Where the Automated Content Engine creates content from a knowledge base, this engine repurposes finished work. The input is content that’s already done and proven; the output is distribution assets (posters + excerpts) plus an enriched original article. Each stage runs as its own scheduled job — observed as daily runs: Blog viz daily, Blog viz poster daily, Blog viz insert daily.

The pipeline

A finished blog/series is handed to a chain of single-purpose agents:

  1. Analyze / read — agent is given access to the finished blogs and reads them to extract the postable ideas, quotes, and hooks.
  2. Brief — agent writes a content brief for each extracted idea (the angle + what the visual should say).
  3. Poster — agent generates a poster from that brief (the visual asset).
  4. Insert — agent inserts the poster back into the article — enriching the original long-form with native visuals.
  5. Post — agent publishes the poster with an excerpt drawn from the article, driving back to the source.
finished blog/series ─► read & extract ─► brief ─► poster ─► insert poster into article
                                                              └─► post poster + article excerpt ─► social
   (already done)         (Blog viz)      (brief)   (Blog viz   (Blog viz insert)
                                                      poster)

Why this is high-leverage

  • Compounds work you’ve already paid for. The expensive part (the writing) is done; this engine extracts near-unlimited distribution from a fixed corpus — the core “Naval-bot” insight: one good essay → hundreds of posts.
  • Two-way value. It doesn’t just spit out social posts; inserting posters back into the article makes the original better (more visual, more shareable).
  • Single-purpose agents = clean handoffs. Each step (read, brief, poster, insert, post) is a small, delegatable interface — easy to debug, swap, or improve in isolation. See Part A2 - Multi-Agent Orchestration.
  • Excerpt-as-hook drives traffic back to the source, so distribution feeds the owned asset rather than replacing it.

Relationship to the other engine

Content EngineBlog Repurposing Engine (this)
SourceProduct/market knowledge baseFinished blogs/series we own
ModeCreate net-new anglesRepurpose proven long-form
OutputDaily TikTok/IG postsPosters + enriched articles + excerpt posts
Mental modelResearch desk”Naval Ravikant bot”

Both should share the same three governing policy layers (channel / brand / performance) — see The three governing policies.

Status

Built and running (daily jobs: Blog viz / poster / insert). Reference implementation for repurposing owned long-form into Distribution assets.