AI Visibility Tool: See how AI search ranks your brand.

Connect AIsa to your agent to check brand mentions, compare competitors and inspect cited sources. Turn the findings into an AI visibility report you can share.

Connect your agent

1

Paste one prompt into Codex

Paste this into Codex to add the AIsa MCP server and start browser authorization.

Please install the AIsa MCP server for me, then confirm it connected.

It's a remote (streamable HTTP) MCP server.
MCP Server URL: https://mcp.aisa.one/mcp

AIsa uses OAuth — open the browser sign-in when prompted. It's read-only public data, so there's no API key to paste.

2
Prefer to set it up by hand?
Add AIsa to ~/.codex/config.toml:
```
[mcp_servers.aisa]
url = "https://mcp.aisa.one/mcp"
```
Then sign in with OAuth:
```
codex mcp login aisa
```
3
Create your own prompt
Go to the prompt builder below, customize your task, then copy the finished prompt into Codex.
Customize my prompt

Does AI mention your brand?
Find out in your report.

Your agent + AIsaDemo · fictional data

You

Create your visibility report.
Start with your brand.

Set the scope of your report

Replace the bracketed fields, then copy the prompt into Codex with AIsa connected. Add your own customer questions or ask the agent to propose a set before running the research.

Brand and website

Market, language and budget

Competitors (optional, one per line)

Customer questions and report goals

Your agent runs the research after you approve its scope and estimated cost.

View full report prompt

Create an AI visibility report using available AIsa tools.
Brand and website: [brand] · [website]
Competitors: [competitor brands and domains, or ask the agent to suggest them]
Market, language and budget: [country] · [language] · [maximum research budget]
Questions and goals: [questions buyers ask before choosing a product or service]
[who will read the report and what they need to decide]

Before paid calls, propose the question set, engines and operations, verify availability and pricing, estimate cost and wait for my approval. Ask for a budget if missing.
1. Test a defined set of buyer questions on available answer engines. Record provider, engine, date, market and raw responses. Label Anthropic Web Search and OpenAI Web Search as API sources if used.
2. Compare brand mentions and website citations with competitors. Show counts and denominators; separate failed calls from valid results. Do not present the sample as overall market share.
3. Inspect cited sources. Add Similarweb, Semrush or DataForSEO research where useful and within budget. Keep traffic estimates, SEO metrics and AI observations distinct.
4. Recommend five prioritized actions with evidence, limitations and follow-up tests. Assemble an executive summary, method, question-level results, citations, action plan and appendix. Provide a structured document and export a PDF if your environment supports it.
Do not reuse demo figures or invent results. Mark missing data, do not equate API responses with all user experiences, and do not promise rankings. Write in English. Do not publish or contact anyone.

Set your report budget.

Start with pay-as-you-go access. The cost depends on the questions, engines and supporting research. Ask for an estimate before running; model and agent fees may apply separately.

Compare all plans

Your AI visibility tool:
your questions answered.

How does this AI visibility tool work?

Connect AIsa to your agent and send a research prompt. Your agent records whether a brand appears in answers to a defined set of customer questions, which competitors appear and which sources are cited. A useful report includes the method, date, missing results and evidence behind each recommendation.

Which AI answers can I compare?

AIsa offers tools such as Oxylabs for supported answer engines, plus Anthropic Web Search and OpenAI Web Search. Ask your agent to verify current availability and label each source accurately. API responses are not interchangeable with personalized answers in consumer apps.

How is visibility measured?

Define your questions, market and engines first. For example, count brand mentions among successful responses and show the numerator and denominator. Report failed queries separately. This is a snapshot of your sample, not total market share or a guaranteed ranking.

Can I export a PDF report?

Ask an agent with file-generation capabilities to prepare a readable report and export it to PDF. Include the supporting tables and source links. File creation depends on your agent; this page prepares a prompt and does not generate a PDF itself.

How does SEO support the report?

Similarweb can add traffic estimates; Semrush and DataForSEO can add search and keyword context. Keep these separate from AI mentions, then use the combined evidence to prioritize pages and content to test. SEO metrics alone do not establish AI visibility.

Can I use the report with clients?

Yes. Define the scope with your client and review the evidence before delivery. The tutorial also explores prospect research; it is optional and separate from this reporting workflow. Reports cannot guarantee future citations or sales.

Is this a live visibility scan?

No. The demonstration uses fictional findings. Your connected agent performs fresh research when you submit your prompt and approve the budget. Repeated monitoring requires a separate schedule in a supported environment.