01 · Work · ScribbleFlow
A terminal tool that takes one topic and returns a finished, SEO-shaped markdown post. You give it a topic, tone, audience and word count; it does the research, outline, drafting and SEO pass on its own.
The drafting runs on a local LLM through Ollama, so there is no API key and your drafts never leave the machine. It is built as swappable parts: research, writing and prompts are separate files, and changing the model is a one-line edit.
02 · Story
Why I built it
I wanted research and a first draft for a post without paying per word or handing my drafts to a cloud service.
Who stays in controlRuns offline on a local model, with no API key.
03 · How it works
Enter a topic, tone, audience and word count at the terminal.
It searches DuckDuckGo and scrapes the results with BeautifulSoup, no API key.
It builds an outline and drafts each section with a local LLM running on Ollama.
The draft gets a keyword and structure pass, with prompts kept in a separate model-agnostic layer.
The finished post is written out as a clean markdown file.
04 · Features
It takes the top five DuckDuckGo results, follows the redirect links to the real pages, and scrapes every paragraph.
The model condenses the sources into a summary, trends, facts and angles before a single section is written.
An SEO outline gives every section a word budget, and each section is drafted in its own call so long posts stay on track.
A hook, an SEO title and a meta description of 160 characters or less, aimed at a grade 8 to 10 reading level.
Runs on Ollama with qwen2.5:3b by default; switching models is a one-line change.
05 · Scenarios
A solo developer
A researched first draft, like the sample post on prompt injection, without a paid API.
A small company blog
Researched drafts on a schedule, without paying per word or sending drafts to a cloud service.
A model tinkerer
The same brief run through different models by changing one line.
I am not looking for work. I am looking for good people to know.