Purpose-built agents for the work you actually need
Choose an agent for research, strategy or content creation, or start one from a Cited task with the relevant inputs already filled in.
Turn AI search data into finished work.
Ask for a brief, report, analysis or spreadsheet and get a finished deliverable built from the AI search data already in Cited. No copy-pasting dashboards. No starting from a blank page.
Not a blank chat window. An agent starts with your Cited data, a defined workflow and a specific output.
The Cited dashboard, running on sample data.
Grounded in the data your team already trusts
Agents read the same Cited data as your dashboards, so the numbers in the deliverable come from the same underlying source.
Fixed, traceable steps
See how the agent reached the result. Each agent follows a defined workflow, so you can inspect the steps and the inputs behind every run.
Every run leaves a reusable deliverable
Every run is saved, and versioned where the workflow supports it, so you can export the result, come back to it later, or run the workflow again when the data moves.
Questions about Agents
Still deciding? Talk to us or read the guides.
Can I run an agent today?
Not yet. The workflows are built and running against production Cited data internally, while the run experience customers will use is still being designed, so there is nothing to switch on in your account. Availability is limited to early brands. Talk to us if you want to be one of them.
How is this different from MCP?
MCP is the reading side. Your assistant asks Cited questions in whatever tool you already work in, gets numbers back, and cannot change anything. Agents are the doing side. They run inside Cited, take a piece of work from start to finish, and save what they produce. Every run is metered and logged against your brand.
Which plan will include agents?
The intention is Scale and above, with a fixed credit price per agent drawing against your plan rather than adding a bill. Nothing is settled until the feature opens, and the pricing page will carry the availability and the allowances when it does.
Will an agent change my site?
No. Agents produce analysis, recommendations and deliverables. An agent writes a draft into your repository and nothing gets published, pushed into your CMS or sent anywhere on your behalf. Deciding a piece of work is finished stays with a person.
Where does an agent get its data?
From the Cited data available to that workflow: your tracked AI search measurements and the supporting evidence behind them. Which inputs an agent reads depends on the agent, and the run shows you which ones it used.
What can I do with the output?
It depends on the agent. Outputs include reports, briefs, analyses and spreadsheets, and each one is saved so you can export it, come back to it, or run the workflow again when the underlying data has moved.
Three kinds of work.
The library is grouped the way the work is, so you pick an agent by what you are trying to produce.
- Research
- Understand the sources, competitors, citation patterns and gaps in your category.
- Strategy
- Turn those findings into a structured recommendation or a plan.
- Create
- Turn the analysis into a brief, report, sheet or other usable output.
Give an agent the goal. Get the work back.
An agent does more than generate text. It follows a defined workflow, reads the Cited data it needs, gathers the evidence behind it, and produces the output in a format your team can use. Task OS tells your team what needs doing; Agents take a supported task and turn its inputs into the deliverable.
- Goal
- Research
- Analyse
- Compose
- Deliver
Agents is one part of the loop.
Cited finds the visibility gap. The Task Operating System turns it into prioritised work. Agents turn supported work into finished deliverables.
Source IntelligenceFind the sources behind the answer.AI does not recommend brands out of nowhere. It reads pages, publishers, communities, marketplaces and other sources. Source Intelligence shows you which sources shape your category, where your competitors are getting cited, and where you are missing.OpenTopic IntelligenceSee where you win and lose, topic by topic.Buyers do not ask AI about your brand. They ask about what they need. Topic Intelligence shows where you stand on the subjects they ask about, which competitors are ahead, and how your position changes by AI platform.OpenAI ShelfSee the products AI puts on the shelf.AI does not just recommend brands. It can put specific products in front of the buyer. AI Shelf shows which products get listed, how often they appear, where they rank, and which competitors are taking the shelf.OpenTask Operating SystemTurn every AI visibility gap into work.Cited turns AI search findings into a ranked list of actions. Every task tells your team what to change, where to change it, why it matters, and what to do next, with the evidence behind the recommendation.Open
Turn your Cited data into finished work.
Free report across three AI platforms. No credit card required.