From the market radar: rising topics the keyword pipeline cannot see

Keyword data describes demand that already has a name. A topic that started moving this week has no search volume yet, so a keyword pipeline cannot rank it. The radar covers that gap. It reads recent discourse about your niche and returns the narratives that are rising, so your agent can brief a timely piece while the topic is still moving.

The radar is not rank tracking and not a news alert. It ranks what people are discussing by how fast it moves.

What a radar snapshot is#

A snapshot is one reading of your niche at one moment. VarynForge derives search queries from the niche, collects recent items from Google News, Reddit and Hacker News, and groups them into ranked rising topics. Google Trends adds a trajectory for the niche's head terms, which orients the reading but does not score the topics. Each topic carries:

• label and summary: what the narrative is.
• momentum: breakout, rising or building.
• novelty: new, partially_covered or covered, judged against the clusters your project already holds.
• drivers: why it is moving.
• evidence: the items behind it, each with its source and title, and a link when one could be matched.

When the radar is lit#

Every new project gets one preview snapshot when it is created, on any plan. The preview is a first look and does not refresh.

A completed research run lights the radar for 30 days. While it is lit, the radar takes a fresh snapshot about once a day. A second run resets the window to 30 days from its completion; the days do not stack. When the window ends, the radar goes dormant: it takes a reading once a week, and everything you already have keeps working. A research run is paid, so Free and paid covers what one costs.

Read the topics: list_radar_topics#

Your agent calls list_radar_topics with the project id. It returns the latest snapshot: its signalObservationId, when it was captured, whether it is a preview, when the window ends, and the rising topics. Each topic carries an index and an addedCount, how many suggestions already came from it. It returns null when the project has no snapshot yet.

Turn a topic into angles: expand_radar_topic#

{ "projectId": "…", "signalObservationId": "…", "topicIndex": 2 }

expand_radar_topic returns three to six candidate article angles for one topic, each with a title, a search intent and an editorial positioning: neutral, contest, agitate, twist or ride_the_wave. Each angle is checked against your content plan. An exact title match is marked already added, and a close match names the suggestion it duplicates. Pass customPov with a point of view in a sentence to get extra angles written from it. The call adds nothing to your plan, so your agent can show you the angles and let you choose.

Commit the angles: add_radar_topic#

add_radar_topic takes the same signalObservationId and topicIndex plus the angles you chose. It adds each one to the content plan as a radar-born suggestion, keeps the topic as its provenance, skips exact-title duplicates and creates the content plan if the project has none. It returns the new suggestionIds.

Each suggestion starts its brief on its own, so your agent does not call generate_article_brief after it. The brief is written from the narrative, its drivers and evidence, in the stance of the angle. On the free plan these briefs count toward the 10 a day, and any past the cap queue and forge when a slot frees. Your agent calls download_brief_markdown once a suggestion reports brief_ready.

When to prefer this path#

Use the radar when you want a timely piece about something your niche is discussing now. Use your own ideas when you already know what to write. Use the data-backed path when you want pieces ranked against search demand. The three paths end at the same brief, and from there drafting and verification is the same.

The Radar screen in the app#

The web app has a Radar screen that shows the same topics as a card deck across your organization's projects, one topic at a time, momentum first. Keep a card to add its chosen angles, dismiss it, or put it back for later. Cards from a dormant or preview radar stay usable and say so.