The Queries screen: every keyword a research run found

The Queries screen sits under Queries in the Data Explorer part of each project's sidebar. It is the raw search demand behind the project: every keyword a research run discovered for the niche. Opportunities groups these keywords into clusters; Queries shows them one by one.

What the list shows#

A table with one row per keyword:

• Query: the keyword text, linked to its detail page.
• Intent: informational, commercial, transactional or navigational.
• Volume: a demand bucket, one of High, Average, Long tail or No traffic. VarynForge sizes demand in buckets, not as a searches-per-month figure.
• KD: keyword difficulty, from 0 to 100, colored by how hard it is.
• Updated: how long ago the project's latest research run completed.

The list starts with the easiest keywords. Keywords with no difficulty score sort to the end, so long-tail noise with an unscored 0 does not crowd the top. It loads 50 at a time.

The keyword detail#

Open a keyword and its page, at /queries/<queryHash>, shows the keyword as the title with its intent, KD and demand bucket. Two sections follow:

• Related Keywords: the keywords stored as related to this one. A related keyword VarynForge holds data on links to its own detail page; the rest show as plain tags.
• Ranking Pages: the pages that rank for the keyword, in position order, with title, URL and company. A page on your own site is marked Yours, and its company reads Your site. Each page links to its page dossier and each company to its company detail.

Where the data comes from#

Keywords, intent, demand bucket and difficulty are written by research runs. A run discovers the keywords for the niche and reads the search results for each one, which is where the ranking pages and their positions come from. The screen has no data until a project's first run, and says so: run a research to map search demand in your niche.

The data is as fresh as the latest completed run. A note under the title says how old it is. After 30 days it adds a Stale badge, after 90 an Antique badge, each with a Re-run link to the project overview. Nothing on this screen refreshes between runs.

What you can do here#

• Intent buttons filter the list to one intent. Pressing the active one again, or Clear filter, removes it.
• Load more fetches the next 50 keywords.
• Click a keyword to open its detail, then a page or a company to follow it.

The screen is for reading. Nothing on it changes the project's data or scores.

The same thing over MCP#

On the screen	MCP tool
The list	list_keywords
Keyword detail	get_keyword_detail

The screen and the tools make the same call. list_keywords takes a projectId, a limit up to 100, a cursor for the next page, an optional filters.intent, and sorting by volume, difficulty, text or intent. Each item carries queryHash, text, difficulty, intent and volumeBucket (no_traffic, long_tail, average or high). The response's nextCursor is absent on the last page.

get_keyword_detail takes a queryHash and a projectId. It returns the keyword, its ranking pages with position, company and isOwned, and resolvedRelated, the related keywords that have a queryHash to open next. An agent picking easy topics should prefer list_opportunities, which scores whole clusters; see the data-backed path.