Answer · ChatGPT plugins

Can you optimize a ChatGPT plugin for discovery?

Partly. You can test and improve when ChatGPT calls your plugin once it is installed, using a method OpenAI documents, but you cannot request, buy or guarantee a recommendation. OpenAI tells developers to build a labelled set of prompts, replay it, and track how often the right tool runs. That part is measurable work. Suggestions to people who have not installed the plugin are granted by OpenAI, and no published evidence shows that any wording change earns them.

First published 9 Oct 2026Sources checked 9 Oct 2026

Last updated 9 Oct 2026

The evidence

What can you change?

Each point carries the page it came from. Platform behaviour is described in OpenAI's own words where we can.

What OpenAI recommends

Gather a golden prompt set of direct, indirect and negative prompts, replay it with the plugin installed, and track precision (did the right tool run) and recall (did the tool run when it should).

OpenAI, Developer docs, optimize metadata

What you control

The plugin name, its short and long description, its category, optional keywords, and each tool name and description. OpenAI advises changing one field at a time.

OpenAI, Developer docs, optimize metadata

What you do not control

Directory placement and proactive suggestions. OpenAI selects plugins for them and developers cannot request it.

OpenAI, Developer docs, publish and review

Where it stops

Text that tells the model to prefer your plugin, or that interferes with fair discovery, is against the guidelines.

OpenAI, Developer docs, plugin guidelines

Limits

What we do not know

  • OpenAI's guide covers tool selection for installed plugins. It does not say the same changes affect suggestions or directory ranking.
  • We have not run a controlled test of a metadata change. Until one exists, we do not claim any change causes a recommendation.
  • Sometimes the wording is fine and the plugin cannot do the job asked. That is a product gap, and no description fixes it.

Nobody outside OpenAI knows the full recommendation method. We report what is documented, what we counted, and what is still unknown, and we keep the three apart.

Sources

Sources and dates

Source date is the date on the source itself. Checked is the day we last read it.

Sources for this answer, with the source date and the date KailxLabs last checked each
SourcePublisherSource dateChecked
Developer docs, optimize metadataOpenAINot dated9 Oct 2026
Developer docs, publish and reviewOpenAINot dated9 Oct 2026
Developer docs, plugin guidelinesOpenAINot dated9 Oct 2026

Related: Do descriptions affect discovery? · Can a plugin pay for placement? · ChatGPT plugins research hub

Want this measured for your plugin?We are testing the method now. Tell us about your plugin and we will say what we can and cannot count yet.
Book a call
Last updated 9 Oct 2026
  • 9 Oct 2026: answer published, from OpenAI pages read at the source that day.
Book a call

Talk to the founder.

20 minutes with Kailesk Khumar. You leave knowing where AI names you today, and what it would take to become the answer.

  • A live look. We ask five of your buyer questions on ChatGPT, Google and Perplexity while you watch.
  • The reasons. The three biggest reasons AI names a rival instead of you.
  • The plan. A clear next step and its price, whether or not you hire us.

Pick a time that suits you. The calendar invite arrives straight away. Prefer a link? Book on Cal.com.

KK Kailesk Khumar AI visibility strategy call
20 min Video call Your time zone
Prefer email? Get a free written read instead
What you sell
How did you find us?

No newsletter. A reply from a person within two working days.

Got it. Kailesk will reply within two working days.
Talk to the founderBook a call