WorkerKitAI workers for Exa
Ready-made AI workers that search by meaning with Exa: find pages like this one, discover companies that resemble your best customers, and get results keywords would never have reached.
- Link Opportunity Finder
Finds the sites that link to your competitors but not to you, ranks them, and builds your prospect list.
- Competitor Newsletter Tracker
Reads competitor newsletters and archive pages, and only briefs Grok Bot when a High product, pricing, or GTM signal appears.
- Regulatory Pulse
Tracks SEC and industry rule drops for your beat, scrapes primary texts when needed, and hands Grok Bot a sourced compliance pulse.
- Changelog Radar
Watches competitor changelogs and release notes and only surfaces High customer-impacting releases.
- Hiring Signal Monitor
Watches competitor and peer hiring signals — career pages, job posts, and "we're hiring" news — and briefs Grok Bot on what roles are heating up.
- Earnings Call Tracker
Tracks upcoming and fresh earnings calls for your watchlist, pulls key takeaways from IR pages and news, and hands Grok Bot a cold briefing.
- Launch and Feature Watch
Scans news and launches for your competitor watchlist and only briefs when a High launch or feature drop is evidenced.
- Funding and Deal Radar
Scans overnight funding, M&A, and strategic partnership news for your sector watchlist and ranks what Grok Bot should surface.
What is an AI worker for Exa?
An AI worker using Exa is an AI agent that can search the web by meaning rather than by keyword, which suits the jobs where you know what you are looking for but cannot name it in the words the page would use. It is a lookup step inside a larger job, and the worker carries what it finds into whatever it was doing.
What these workers do
Find companies like your best customers
Takes an example customer and retrieves organisations of similar character, which is prospecting done by resemblance rather than by industry code.
Discover sources on an unfamiliar topic
Starts from a description of what is needed and returns material that matches the idea, including pages that never use your phrasing.
Find more like this page
Given one useful source, retrieves others of the same kind, which is how a shallow search becomes a real reading list.
Map who else is working in a space
Surfaces the organisations and writing around a topic so a market picture is assembled from what exists rather than from memory.
How an AI worker runs in Exa
Describe it, do not name it
Find similar is the distinctive move
When it is the wrong tool
Exa questions, answered.
How is meaning based search different from keyword search?
Keyword search matches the words on the page, so you must guess the vocabulary the author used. Meaning based search matches the idea, so a description or an example page can retrieve relevant results that share no wording with your query at all.
When should a worker use this instead of an ordinary search?
When the request is a shape rather than a phrase: find things like this, find who else works this way. If you already know the exact term, a keyword connection is faster and more precise, and a well written kit chooses accordingly.
Does it return sources I can check?
Yes. Results come back as pages the worker names in its output, so a finding can be followed to its origin. That matters more here than usual, since results reached by resemblance deserve a look before they are acted on.
Can it search inside my own documents?
No. This is a web connection. Your own documents are reached through the connections you grant for those systems, and keeping the two separate is deliberate: internal knowledge and public discovery are different permissions.
What AI workers can do in Exa
The tools a worker can be granted in Exa once you connect it. A kit grants only the ones its job needs, and you can narrow them further on the worker.
Exa Web Search
Search the public web with Exa and get ranked results (title, url, snippet, and — where visible — content, score, published date, author, source). Set `includeContent` to pull page text inline where supported. Pass startPublishedDate/endPublishedDate to pin a published-date window. Read the `accessInfo` field: results are filtered by the connection's egress (domain/URL allow-block) rules, so a short list may be scoped rather than exhaustive.
Exa News Search
Search recent news across the public web with Exa, returning ranked news results (same result shape as exa_search). Pass startPublishedDate/endPublishedDate to pin the window for dated questions (e.g. yesterday's news) — without them the engine's default recency is NOT a specific window. Read `accessInfo` for any egress filtering applied.
Exa Read Page Contents
Fetch clean, readable page text for one or more URLs with Exa (e.g. to read a page from exa_search). Each URL is checked against the connection's egress rules before fetching; blocked URLs are dropped and summarised in `accessInfo`.
Exa Answer
Ask Exa a question and get a synthesized answer with citations, returned in `answer` alongside the supporting `results`. Cited sources are subject to the connection's egress rules — read `accessInfo` for any filtering.
Exa Find Similar Pages
Given a URL, find semantically similar pages on the public web with Exa. Results are subject to the connection's egress rules — read `accessInfo` for any filtering.