← All help topics

Overview

The home screen shows the project's pulse: how many posts await approval and what's already published.

1Metrics

Cards show drafts, approved and published posts, and active sources.

2Project dashboard

The Overview is grouped by meaning so it's clear at a glance. At the top — a plain-language summary line: what's happening with the project right now (how many drafts await approval and when the next publication is). Then blocks:

  • 📝 Your posts — drafts / in plan / published and the latest posts.
  • 🔌 Connections — topic sources, bots and integrations, publishing channels.
  • 🔄 Your AI content-marketing pipeline — a visual data-flow diagram.
  • ⏱️ A countdown bar to the next automatic search.
  • 💳 Usage and limits — weekly tokens, daily chart, generation limit and the number of model requests.

Each tile links to its section.

3🔄 Your AI content-marketing pipeline

Below the cards is an animated pipeline diagram where each entity is a separate node:

  • 🔗 every active source → 🧠 models → 🗓️ post plan;
  • if auto-posting is off — a “You” node for approval;
  • → 🤖 every bot/integration → 📣 every channel.

The connection lines are animated — a running data flow. What you can do:

Sources are ordered by effectiveness: the more productive a source, the further right it sits (closer to “Models”); each source node shows three effectiveness dots.

🔬 The “Show processing” button expands, from the “Models” node, the information-processing mechanics as a sub-chain of stages: 📥 collect material (parsing + full text) → 🧭 selection (relevance and dedup) → 🧠 brand rewrite → 🔬 fact-check (facts, date, numbers, anti-self-promotion, opinion scrutiny) → 🖼 image. Hover a stage to see exactly which checks run there (and, when available, the numbers: how many items were fetched, filtered, rejected, and posts created over the recent sweeps). This makes the whole “material → finished post” path obvious.

Lines from sources to “Models” are colored by source effectiveness (like the dots on the chip): 🟢 green — effective, 🔴 red — ineffective, ⚪ grey — effectiveness still being computed; line thickness reflects how many posts came from the source. A source node also shows numbers: 📝 posts and the “posts per 1k tokens” efficiency.

Below the sources, 📋 post templates are shown (if any) — they feed the models too; a template tooltip shows the schedule, next and previous post, channel, type and status.

When the mechanics are expanded, there's a funnel-period switch nearby: “per sweep / 7 days / 30 days”.

More buttons: “🗺 Map” shows/hides the minimap (hidden by default); “⛶ Fullscreen” expands the graph to the whole area (close with the button or Esc); “⊟ Collapse all blocks” folds sources, bots and channels into compact aggregates (and back with “⊞ Expand all blocks”). The “Post plan” node tooltip shows post counts: drafts in queue, approved, published. The dashboard header has a 🔀 (AI pipeline) icon button — it opens the graph fullscreen with blocks and mechanics expanded from any section. The graph data refreshes automatically about once a minute (with up-to-date stats and sorting).

The layout is built automatically left-to-right.

4Generate drafts

The button runs the AI pipeline: it scans sources and queues new drafts.

5Start / Pause the project

At the top-right of the Overview there's a “⏸ Pause” / “▶ Start” button, and next to the name and in the menu (at the project switcher) you can see the status: 🟢 active or 🔴 paused. A paused project does NO background work: no source sweep, no trend search, no template generation, no auto-publishing. Handy to temporarily “freeze” a project without deleting it or spending tokens.

6A new project starts paused (auto-start)

A freshly created project is paused immediately — so background processing doesn't waste resources on a project that isn't set up yet. As soon as the project has all three pieces — a post plan, a channel and a bot — it starts automatically (you'll get a “Project started” notification). After that you're in control: the “⏸ Pause / ▶ Start” button works as usual, and once you've toggled it manually the auto-start no longer interferes. So until a project is fully configured it doesn't spend tokens.

Didn't find an answer? Write to us — we'll help.

Contact support