Skip to main content
Plans: All plans, limits vary · In the app: Sidebar → Agent → Tables
A table template is a working table in one click: the agent is attached, the input, agent and output columns are set, and up to three sample rows are filled from your own workspace (tracked prompts, topics, cited pages, competitors’ cited pages, your domain). Where your data runs short, the sample rows use written examples with your brand filled in. The table is padded to 100 rows, ready for your own inputs.

Use a template

1

Open the gallery

On the Tables home, click Template under Create a new table. Search by name, description or category.
2

Pick one

Click a card. The table is created immediately and named after the template.
3

Add your rows and run

Replace or add rows, then click ▶ Run on the agent column. See Tables.
Each template also adds a published copy of its agent to your Agents list, marked “From a table template”. Edit that agent in the builder like any other; the table follows its latest version unless you pin one. A template whose agent needs something your workspace does not have yet stays in the gallery, greyed out, with the reason.

Templates

Content creation

Content brief

One row per keyword. Each gets a brief built from the web and your tracked AI answers: title options, an outline, the questions to answer and what competitors already cover. Output: Final brief (Long Text).

Articles from briefs

One row per brief. Each becomes a full draft written to the house rules, then polished against the rules on machine-made writing. Outputs: Article (Long Text) and What the polish changed (JSON).
Chain it: run Content brief, then point this template’s Brief input at the Final brief column.

Answer-first openings

One row per question you track. Each gets a researched 40 to 60 word opening an AI engine can lift as the answer. Outputs: Opening answer (Long Text) and Claim (Text).

Article translator

One row per article and language. Each is read, cleaned, translated and localised with its headings, links and SEO structure kept, then checked by a second pass. Sample languages: Spanish, German, Japanese. Output: Translated article (Long Text).

Content optimization

Content refresher

One row per article of yours. Each is checked against what has changed since it was written and comes back refreshed, with a list of what changed and why. Outputs: Refreshed article and What changed (Long Text).

FAQ generator

One row per page. Each gets the FAQ section it should end with, from the questions you track, and the FAQPage structured data to publish with it. Outputs: FAQ section (Long Text) and FAQ schema (JSON). One row per page. Each is read against your sitemap and gets the internal links to add: the anchor text, the page it points to, where it goes and why. Outputs: Link suggestions (Long Text) and Pages considered (Number).

Product page optimization

One row per product page. Each gets a rewritten title, description and spec table that AI shopping answers can quote, with what changed and why. Outputs: Optimised title (Text), Optimised description, Specs and attributes, What changed and why (Long Text).

Page citation audits

One row per page of yours. Each is read, checked against the answers that cite it, and given the five edits that would make it the citable answer. Outputs: Audit (Long Text) and Citations today (Number).
Fill it fast with Import → Cited pages: your pages that AI engines already cite, most cited first.

Research and reporting

Prompt ideas by topic

One row per topic. Each proposes ten new questions worth tracking, from what you already track and what people search around it. Output: Proposed prompts (Long Text).

Topic news digests

One row per topic. Each becomes a five-line digest of this week’s news with a source under every line. Output: Digest (Long Text).

Outreach

Citation outreach

One row per page the engines cite instead of yours. Each gets the email asking its author to reference you, with the specific reason. Output: Email (Long Text).
The email is written into the table. Nothing is sent. Copy it into your own mailbox to send it.

Next steps

Tables

Import, run, review and export.

Routines

Recurring reports Quotable Agent writes on its own.

Column types

What each column holds.
Last modified on November 18, 2026