@joemccann/pi-exa
v1.0.0
Published
Exa AI semantic search and similarity tools for pi coding agent — search the web by meaning, find similar pages, filter by domain/date/category
Maintainers
Readme
pi-exa
Exa AI semantic search, similarity discovery, and content extraction for pi.
Install
pi install npm:@joemccann/pi-exaOr try without installing:
pi -e npm:@joemccann/pi-exaSetup
Set the EXA_API_KEY environment variable. Get a key at dashboard.exa.ai.
export EXA_API_KEY=your-key-hereTools
| Tool | Description |
|------|-------------|
| exa_search | Semantic web search — find pages by meaning, not keywords |
| exa_find_similar | Find pages similar to a URL — competitors, related content |
| exa_get_contents | Extract clean text from URLs — more reliable than scraping |
exa_search
Semantic search with filtering by domain, date, category, and search mode.
Search types:
auto— Default, picks the best modefast— Sub-350ms responsesneural— Embedding-based semantic searchdeep— Multi-step reasoning for complex questions
Categories: company, research paper, news, tweet, personal site, financial report
Content options: includeText, includeSummary, includeHighlights (default: true)
exa_find_similar
Finds pages semantically similar to a given URL. Pass a company site to find competitors, an article to find related content, or a product page to find alternatives.
exa_get_contents
Extracts clean, parsed content from one or more URLs in a single API call. Returns text, highlights, and/or AI summaries.
Examples
> Search for recent AI safety research papers
> Find companies similar to Stripe
> Extract the main content from these three URLsDevelopment
git clone https://github.com/joemccann/pi-exa
cd pi-exa
npm testTest locally with pi
pi -e /path/to/pi-exaLicense
MIT
