CtrlK
BlogDocsLog inGet started
Tessl Logo

bapfernandez/article-creator

Content creator for tessl.io — generates publish-ready blog articles with SEO metadata, Tessl house style, and technical authority.

90

1.26x
Quality

79%

Does it follow best practices?

Impact

92%

1.26x

Average score across 10 eval scenarios

SecuritybySnyk

Passed

No known issues

Overview
Quality
Evals
Security
Files

task.mdevals/scenario-4/

Task: Thought Leadership Article on Context Engineering

Background

The Tessl content team is preparing a new push on the topic of context engineering for AI agents. The content lead has asked for a publish-ready thought leadership article that can anchor a broader editorial series. The target audience is senior engineers and technical leads who are already building with LLMs and are starting to think about production reliability.

The content lead's brief notes:

"We want something that challenges how engineers currently think about this problem. Not a tutorial, not a product pitch. Something that makes a smart engineer pause and say 'huh, I hadn't thought about it that way.' It should introduce a mental model they can actually use. We need this ready to publish directly on the blog — full metadata, slug, everything."

What to Write

Write a thought leadership article about context engineering for AI agents. The article should:

  • Address a real tension or misconception that engineers commonly hold when they start building with LLMs in production
  • Introduce a named mental model or framework that helps readers think about the problem differently
  • Include at least one concrete, specific example that grounds the framework
  • Be appropriate for the tessl.io engineering blog

The article should be publish-ready, including all metadata needed to put it live.

Output Specification

Save the completed article as article.md in the current working directory.

The file must include a metadata block at the top (title, type, primary keyword, meta description, URL slug, internal links, estimated read time) followed by the full article body in markdown.

evals

tile.json