Visitors
Who brings visitors?
See attributable visits from AI assistants and their landing pages. Find out where AI traffic reaches your website.
AI analytics for your website · Free pilot
What happens after the AI click?
See which AI assistants send visitors, which tool calls fail, and which inquiries your server confirms. Record the problem, document the fix, and test the path again.
Insights brings observed issues and next steps together. Your site history keeps the changes and their actual test results.
Open source. Hosted in Germany. No credit card. See the demo page first if you like.
What do your ChatGPT tool calls actually produce?
Connect your MCP server to AgentTracking. See calls, errors and duration by plugin version, alongside server-confirmed PDF creation, PDF delivery, inquiries and bookings. Test runs and production have separate views.
AI agent analytics for websites
AI traffic is the beginning. What happens next?
Agent Tracking helps website teams inspect AI referrals and tool failures, confirm inquiries through server receipts, and document fixes with browser retests.
Three questions worth opening a dashboard for.
Use the evidence to decide what to investigate next.
Tool failures
What breaks on your site?
Inspect observed tool calls, technical errors, and response times. Insights links observed failures to the relevant evidence and a next step.
Results
Did the inquiry arrive?
Your server confirms successfully created inquiries or bookings. Missing confirmation remains unknown, rather than becoming a lost sale.
From a failure to a checked fix.
Keep the problem, the change, and the retest together.
1. Investigate
Start with evidence
Insights highlights measurement gaps, observed failures, and pending retests. Open the supporting data before deciding what to change.
2. Fix
Document the correction
Record a finding, assign an owner, and describe the change. Link the correction to its test runs.
3. Retest
Check the same path again
Repeat a prepared browser inquiry test on your staging website. Compare the outcome and test conditions.
4. Share
Show what changed
Keep checked fixes in your site history and share findings, corrections, and retests in a protected report.
The current retest checks a prepared browser form flow. It does not measure independent AI task completion or revenue impact.
For teams that can change the website.
A practical workspace for investigating and checking improvements.
Agencies
Evidence for each client
Document issues and checked fixes for each client website. Share the evidence with the people responsible.
Website and product teams
Investigate broken tool calls
Inspect technical failures and link inquiries to confirmations from your own system.
GEO and AEO teams
See what reaches your website
Add observed visits and website activity to your visibility analysis. Crawler requests remain separate from citations and recommendations.
Start with the signal you need.
Visitors from AI assistants
Install the snippet for recognized referrals and supported browser tool activity.
Attribute ChatGPT traffic →Crawler access
Connect origin logs to inspect access attempts and confirmed HTML fetches separately.
Check crawler access →Completed inquiries
Add server receipts for successfully created inquiries or bookings.
Set up server confirmation →A sample of the measurement views
Overview, Agents, Tools and Pages show the observed activity. The private workspace also has task tests, fixes and reports. These screenshots use demo data; they are not customer results.
Connect your first data source
Add your site in the dashboard, put this on every page, and press verify. Tools you register through navigator.modelContext are picked up without any change to your code; declarative tools are forms with a toolname.
<script defer data-domain="example.com" src="https://agenttracking.co/agent.js"></script>Documented data selection
The snippet sets no cookies or local storage entries. Raw network addresses and tool argument values are not stored in event records. Daily salted session hashes remain estimates and may require a privacy assessment. Raw events are deleted after 90 days; daily totals stay as long as the site does. Everything lives on our own server in Germany, and the data processing agreement is concluded the moment you add a site. Removing the site deletes all of it.
Cloud or your own server
The whole product is open source under AGPL-3.0: the snippet, the ingest, the dashboard, the log import, the cron jobs. Run it on your own machine with one compose file, or use the hosted version here and let us run it.
Cloud · Free
Free during the pilot
1 site, 10,000 agent events a month, 30 days of history. Plain page views are not counted. No card.
Cloud · Pro and Agency
Not open yet
5 sites and 500,000 events with manifest alerts, or unlimited sites with a white-label badge. Prices follow when the product has earned them; until then write to us and we switch your account by hand.
Self-hosted · Free
Your server, your data
Every plan limit is yours to set. One Node process, one SQLite file, no external service except a mail sender. Source and instructions on GitHub.
git clone https://github.com/shufflethis/agent-tracking.git && cd agent-tracking cp .env.example .env && docker compose up -d
Before you connect your website
What you can measure, what needs an integration, and what the results mean.
Will Agent Tracking improve my rankings in AI answers?
Agent Tracking shows observed activity on your website. You can use it to investigate technical problems and retest changes. It does not guarantee more mentions, citations, or better placement in AI answers.
Can I see what people asked ChatGPT?
Recorded referrals and server logs do not reliably reveal the original question. Closely timed page requests do not identify a specific prompt either.
Is one script enough?
The snippet records recognized referrals and supported browser activity. Crawler analysis requires origin logs. Confirmed business outcomes require a server integration.
What can I check before agents visit my site?
You can run a prepared browser inquiry test on a test or staging website. This synthetic test is shown separately from real usage.
Can I take my data with me?
Statistics are available through API, MCP, and CSV. Protected reports can be exported as JSON. You can also host Agent Tracking yourself.
Building with our first users.
Agent Tracking is new, and the honest way to price something new is to run it first. During the pilot every account is on the Free plan: 1 site, 10,000 agent events a month, 30 days of history. Plain page views are not counted.
Need more than that while the pilot runs? Write to hi@finalmaster.net and we switch your account by hand. Paid plans follow when the product has earned them, and nothing about your account changes without 30 days' notice by email.
What we ask in return: tell us what the dashboard got wrong, which agent it missed, which number you did not believe. That is what the pilot is for.