Skip to content

Prompt Marketplace

Prompt Marketplace gives users a fast way to browse curated prompts without leaving the extension.

Source of truth

The library is stored in Prompt Library Source.

That markdown file is imported into the extension bundle and parsed locally at runtime.

The source now favors richer multi-line prompt templates with placeholders, checklists, and output instructions. The extension keeps the full prompt text for copy/insert/send actions, while popup cards and inline suggestions use compact previews for readability.

Core capabilities

  • browse prompts by category
  • search prompt text locally
  • filter by category
  • show inline prompt suggestions while the user types in supported AI chat inputs
  • copy prompts to clipboard
  • insert prompts into supported AI chat inputs
  • insert and send prompts on supported AI chat pages
  • track local usage for trending prompts

Supported categories

  • Developer
  • Learning
  • Writing
  • Productivity
  • Image Generation
  • Business
  • Daily Life

Marketplace data model

Each parsed prompt is normalized into a local object with:

  • stable id
  • title
  • full prompt text
  • category key and label
  • duplicate count from source import
  • source line numbers
  • derived keywords for search

The extension stores local marketplace usage in chrome.storage.local.

Tracked actions:

  • copy
  • insert
  • send
  • inline suggestion insert

Trending prompts are ranked from:

  • total local usage
  • send weight
  • insert weight
  • copy weight
  • recent usage recency

AI chat integration

Marketplace actions reuse the same AI input detection path as the Quick Prompt Builder.

Supported targets:

  • ChatGPT
  • Claude
  • Gemini
  • Grok
  • DeepSeek

If the active tab is not a supported AI chat page, the marketplace shows a friendly failure message instead of silently doing nothing.

Inline prompt suggestions

AI Dev Coach can surface a lightweight suggestion dropdown directly in the AI chat composer.

The inline suggestions are ranked from:

  • typed query relevance
  • lightweight intent detection
  • role-aware weighting
  • local prompt popularity

The first release stays local-first and reuses the same prompt library stored in docs/prompts.md.