Reports

Six ready-made reports

A report hands your assistant a finished checklist: which data to read, in which order, and how to write it up. You pick the project, the assistant does the rest, and you can ask follow-up questions afterwards.

Two places, the same reports

In your assistant

You start a report as a command. In Claude Code, for example, type /palagus and press Tab to get /mcp__palagus__weekly_pulse. Your assistant then runs the checklist with Palagus’ tools.

In the Palagus app

The same reports run with one click on the Reports page, on plans that include Reports. The topic heatmap is the one report that is not there: it takes too many calls for a report in the app.

In the MCP standard these are called prompts. We say reports, because in Palagus a prompt is one of your tracked questions. New reports reach your assistant when it reconnects; there is nothing to install.

Where your assistant offers them

  • Claude.ai

    Where
    + → Connectors → Add from Palagus
    Then
    The report lands in the message box as an attachment and is not sent yet. Press Send yourself. Type the project's name exactly as it appears in Palagus.

  • Claude Code

    Where
    Type /palagus and press Tab; it becomes /mcp__palagus__weekly_pulse
    Then
    Add the project; it runs right away.

  • ChatGPT

    Where
    ChatGPT does not offer them
    Then
    Ask in plain words instead. The tools are called all the same.

  • Codex CLI

    Where
    Codex does not offer them (checked 2026-10-01)
    Then
    Ask in plain words instead. The tools are called all the same.

The reports

Weekly pulse

weekly_pulse

A short status of one project: where visibility, share of voice and sentiment stand, the biggest movements, which competitors gained ground, and the answers whose sentiment stands out.

Use it when you want a Monday-morning snapshot of what changed.

projectrequired
The project's name, exactly as it appears in Palagus.
rangeoptional
The time window: last1h, last6h, last12h, last24h, last7d, lastMonth, lastQuarter, lastYear or all. Default last7d.

Drop diagnosis

drop_diagnosis

Why a number fell, step by step: the overview, then narrowed down by topic, persona and AI engine, then the sources the answers cite, then the answers themselves.

Use it when visibility or sentiment went down and you want to know where it happened.

projectrequired
The project's name, exactly as it appears in Palagus.
rangeoptional
The time window: last1h, last6h, last12h, last24h, last7d, lastMonth, lastQuarter, lastYear or all. Default last7d.
topicoptional
Narrow the diagnosis to one topic, by its name as it appears in Palagus.

Where you lose

where_you_lose

Where the project trails its competitors (per AI engine and per topic) and to whom, with answers that name the competitor and not you as evidence. Every gap comes with its measurement margin, and a gap within it is called level.

Use it when you want to know where competitors are ahead of you, and by how much it really is.

projectrequired
The project's name, exactly as it appears in Palagus.
rangeoptional
The time window: last1h, last6h, last12h, last24h, last7d, lastMonth, lastQuarter, lastYear or all. Default last7d.
competitoroptional
Compare with this one competitor everywhere. Without it, each engine and topic names its own strongest competitor.

Topic heatmap

topic_heatmapBetaAssistant only, not in the app

Visibility per topic and per AI engine, as a grid: strong and weak spots at a glance.

Why it is a beta: Every cell is a separate call, so the grid is capped at 6 topics × 4 AI engines and can take a while.

Use it when you want to see which topic is weak on which engine.

projectrequired
The project's name, exactly as it appears in Palagus.
rangeoptional
The time window: last1h, last6h, last12h, last24h, last7d, lastMonth, lastQuarter, lastYear or all. Default last7d.

Source authority

source_authorityBeta

The websites AI answers cite for the project (your own against competitors' against third parties) and which engines cite each.

Why it is a beta: It works from citation counts only. There is no rate at which a source is fetched.

Use it when you are working on content and want to know which sites the engines rely on.

projectrequired
The project's name, exactly as it appears in Palagus.
rangeoptional
The time window: last1h, last6h, last12h, last24h, last7d, lastMonth, lastQuarter, lastYear or all. Default last7d.

Campaign tracker

campaign_tracker

Did anything move after a date, such as a campaign, an article or a PR push? It compares an equal window before and after the date, reads the curve across both to tell a change that starts at the date from one already under way, and lists the prompts that moved most. It shows what changed, never that it changed because of the date.

Use it when you launched something and want to see whether AI answers reacted.

projectrequired
The project's name, exactly as it appears in Palagus.
daterequired
The day it started, as YYYY-MM-DD. The two windows meet at midnight UTC of that day.
daysoptional
Length of each window in days, 1 to 90. Default 7.

After a report

The assistant keeps every tool at hand, so you can drill into whatever stands out: one engine, one topic, or the answers behind a number. See the use cases for questions that work well.