AEO for developer tools companies
Get recommended inside the coding assistants and AI answers developers now consult before choosing a library or platform.
Developers adopted AI assistants faster than any other audience, and they use them at exactly the moment of tool selection: which library handles this, what should I use instead of that, how do I do this in that framework. The recommendation arrives inside the editor, and it is frequently acted on immediately.
This is the one category where documentation quality is the whole game. Assistants are trained and grounded on public docs, examples, and the discussions around them. A tool with clear, complete, crawlable documentation gets recommended; one whose docs are behind a login, rendered client-side, or thin on runnable examples effectively does not exist to the model.
Prompts that matter here
- best library for [task] in [language]
- how do I [task] with [framework]
- [tool] vs [tool]
- alternatives to [tool]
What to do
- 1
Make documentation server-rendered and crawlable
AI crawlers generally do not execute JavaScript. Documentation that renders client-side is invisible to them no matter how good it is — a surprisingly common and entirely fixable failure.
- 2
Publish runnable examples for real tasks
Assistants quote code. Complete, copy-pasteable examples organised by task are what get surfaced; API reference tables alone rarely are.
- 3
Write the migration and comparison pages
'Alternatives to X' and 'migrating from X to Y' are high-intent developer queries and are usually only answered by whoever is being migrated away from.
- 4
Publish an llms.txt and keep docs available as markdown
Serving plain markdown alongside the rendered docs gives models a clean, unambiguous source and removes any dependence on their HTML parsing.
Where Elmo fits
Elmo tracks the library and framework recommendation prompts developers actually run, across engines, and reports which sources drove each citation. As an open-source tool it also runs inside the environments developer-tools companies already trust.
Other industries
/ FAQ
Frequently Asked Questions
- How do AI coding assistants decide which libraries to recommend?
- Largely from public documentation, code examples, and community discussion. Tools with thorough, crawlable docs and plentiful runnable examples get recommended far more often than equally capable tools with sparse or client-rendered documentation.
- Does client-side rendered documentation hurt AI visibility?
- Significantly. Most AI crawlers do not execute JavaScript, so client-rendered docs return an effectively empty page. The content exists for human readers and not for the models developers ask for recommendations.
- What is llms.txt and should a developer tool publish one?
- It is a plain-text file that points models at your most important documentation in a clean, parseable form. It is cheap to publish and removes any reliance on a model correctly parsing your rendered HTML.
Ready to track your AI visibility?
Deploy Elmo in minutes and start monitoring how ChatGPT, Claude, and Google AI Overviews talk about your brand. Open source, self-hosted, free.