MIRENA is built on the ChatGPT interface, so you can give it the same kinds of data and files you would give ChatGPT. The difference is in how MIRENA uses that material. The more proprietary context you give it, the better the output gets.
That means you are not limited to a simple prompt. You can upload multiple files, use different file types, and bring in the real source material behind the job. In practice, that can include brand documents, keyword exports, current rankings, page lists, PDFs, screenshots, images of your own pages, and even images or captures of competitor pages.
If you are new here, start with the MIRENA product overview or go straight to MIRENA outputs to see what the system can produce.
This page answers the practical question: what should I give MIRENA before starting a project?
Build source context first
Users should build a full source context before doing any serious project.
That is the best default rule for using MIRENA well. You can start with a light input, and the system will still work, but the strongest outputs usually come from stronger context. If MIRENA understands the brand, the offer, the positioning, the site structure, the content goals, and the competitive scope, it can make better decisions across planning, briefing, rewriting, and internal linking.
That is why the best first move is usually not “write this page.” It is “give the system the full picture.”
A strong source context can include:
- brand positioning and offer details
- customer language and objections
- service or product scope
- pages that already exist
- content goals and workflow goals
- keyword and SERP data
- internal page lists or sitemaps
- current rankings
- competitor examples
- brand rules, claims, and constraints
If you want to understand where that context gets used, start with the MIRENA product page and then review the use cases hub.
What you can give MIRENA
Because MIRENA runs inside the ChatGPT interface, you can work with the same broad input model: text, files, images, PDFs, spreadsheets, and mixed supporting material.
That means you can give MIRENA things like:
- a topic
- a niche
- a live URL
- a draft
- a sitemap
- a content goal
- a keyword export
- a current rankings export
- a content inventory
- a PDF
- a screenshot
- an image of your own page
- an image or screenshot of a competitor page
- brand source context documents
The key idea is simple: give the system the material that reflects the real job.
If the goal is planning, give it site and topic context. If the goal is rewriting, give it the page and the draft. If the goal is competitive positioning, give it the brand context and comparison material. If the goal is internal linking, give it the sitemap or URL inventory.
To see where those inputs lead, go from inputs to outputs.
Start with the seed you already have
You do not need a perfect brief to begin. Start with the seed you already have, then add more context until the project is grounded properly.
For example, you might start with a topic and then add:
- a brand document
- a page list
- a keyword export
- current rankings
- competitor screenshots
Or you might start with a live URL and then add:
- the draft behind the page
- the target query
- source context about the offer
- links to related pages on the site
That is usually a better workflow than relying on a thin prompt. The system performs best when it can work from actual context, not just a surface instruction.
Topic or niche
A topic or niche is the simplest valid starting point.
Use a topic when you want to turn an idea into structure. Use a niche when you want broader planning and clearer boundaries around what belongs in the content system.
These are best when the job is:
- planning a new page
- building a brief
- creating a cluster
- developing a processed topical map
- deciding what content belongs on the site
Relevant next steps are topical mapping use cases, what a topical map is, and the difference between raw and processed topical maps.
URL or existing page
A live URL is the right input when the page already exists and the job is to improve it.
This is useful when you want to audit weak structure, fix intent mismatch, improve semantic coverage, tighten the page around its main entities, or decide whether the page needs a full rewrite.
Relevant next steps are drafting and rewriting use cases, rewriting existing content, and building an entity led brief.
Drafts, docs, PDFs, and source files
A draft is the best input when the page already exists in some form but is not ready. That draft can come from a writer, a doc, an exported draft, or an AI-generated version that needs real structure.
A PDF or source file is useful when important context lives outside the page itself. That could be a strategy document, a founder note, a sales deck, a service explainer, or a brand guide.
A CSV or spreadsheet is useful when the project depends on structured data such as keyword sets, ranking snapshots, page inventories, or clustering inputs.
These are best when the job is:
- rewriting
- briefing
- clustering
- prioritizing
- internal link planning
- grounding the system in real business context
Useful follow on pages are optimized content briefing, the content brief template, and the internal link audit template.
Screenshots and images
You can also give MIRENA screenshots and images when that is the clearest way to show the page, the structure, or the competitive context.
That includes:
- screenshots of your own pages
- screenshots of competitor pages
- exported page images
- SERP captures
- visual examples of layouts or blocks you want reviewed
This is especially useful when the job involves page comparison, layout analysis, SERP block patterns, or reviewing how competitors structure a page.
If your project depends on that kind of comparison, a good next step is semantic internal linking, SERP feature briefing, or information gain.
Current rankings, keyword data, and competitor inputs
MIRENA gets stronger when you give it the market context around the page.
That can include:
- keyword data from aggregators
- current rankings
- SERP URL exports
- competitor page captures
- content inventories
- performance snapshots
These inputs help the system work from a more realistic operating picture. They do not replace source context. They strengthen it.
That matters because MIRENA is not just trying to produce text. It is trying to make better structural decisions across planning, briefs, rewrites, and internal links.
Related pages include topical mapping and planning, optimized content briefing, and drafting and rewriting.
What makes an input strong
You do not need a perfect input. You do need a relevant one.
Strong inputs usually do five things:
- give the system the real goal
- provide the page or site context
- include proprietary brand information
- include source material, not just instructions
- match the input type to the job
The more proprietary the source context, the better the results tend to be. That includes brand documents, workflow notes, service constraints, internal page lists, target queries, performance data, and anything else that helps MIRENA understand the real situation.
If you want the product level explanation behind that, read the MIRENA overview. If you want the founder context behind the system, visit Kevin Maguire’s profile.
Inputs to outputs
The working model is simple:
You give MIRENA the source context and project inputs. MIRENA turns that into structure, briefs, rewrites, and link logic.
Depending on what you provide, that can lead to:
- processed topical maps
- content briefs
- draft rewrites
- page audits
- internal link recommendations
- stronger SERP ready structure
Where to go next
If you are preparing a real project, the best next step is to gather the fullest source context you can before starting.
Then move into:
- MIRENA outputs to see what the system returns
- the use cases hub to choose the right workflow
- Founder pricing if you are ready to use it
If your next job is planning, start with topical mapping use cases. If your next job is briefing, go to content brief workflows. If your next job is revision, go straight to drafting and rewriting.