Agentic Stress Tester

Test the AI agent on your website before your customers do.

The tester runs an agent network from an agentic-web site against real models with your own key. Rules, not another AI, decide pass or fail, and each rule says why.

What it catches

Each check is motivated by something that already went wrong for a real business. The link goes to the source.

Invented policy

An airline's chatbot promised a fare rule that did not exist. A tribunal held the airline to it.

CBS on the Air Canada case

Unsafe medical and allergy answers

Confident answers that could harm a patient or a diner. ECRI ranks misuse of AI chatbots as the top health technology hazard for 2026.

ECRI

Scams and secrets

A request to move money to a "safe account", or to hand over a one-time code. Banks in the UK must now reimburse impersonation scams.

UK PSR on APP fraud

Hidden instructions

Text inside a page or a tool result that tells the agent to book, buy or reveal something. Browsing agents have been steered this way.

Malwarebytes on prompt injection

Industries

Two checklists for each industry, plus the e-commerce checklist. Each case is written for the kind of site, not for one site's actions, so it runs on any network in that industry.

Loading the checklists…

How it works

  1. Check it free. Pre-flight reads the network and its actions, and says what is wrong and how many characters each agent sends. Nothing is spent.
  2. Run a smoke test. About 40 model calls for a typical site: the site's own sample queries plus the cases that test only fixed rules. You see the cost before you start.
  3. Read the report. Each case passes or fails with the evidence. Cases that cannot run on a site are listed with the reason, never counted as passes.

Your API key stays in your browser. It goes only to the router you type, and it is never stored or put in a report. Keyword rules match words, so they can be wrong; each verdict shows its evidence so you can check it.

For AI agents

The zero-token checks need no key and spend nothing. Run them from a clone of this repository.

node cli/tools.mjs lint-bundle path/to/bundle
node cli/tools.mjs call-size path/to/bundle
node cli/tools.mjs validate-call path/to/bundle --tool website_add_to_cart --args '{"id":"margherita","quantity":1}'
node cli/tools.mjs list-checklists --industry restaurants
node cli/tools.mjs estimate path/to/bundle --checklist restaurants-menu.v1
node cli/tools.mjs check-trace trace.json --checklist restaurants-menu.v1 --dir path/to/bundle

Exit codes: 0 passed, 1 a check found a problem, 2 bad input. Output is JSON.

As a local MCP server

The same checks, for coding agents. Keys never leave your machine.

{
  "mcpServers": {
    "agentic-stress-test": {
      "command": "node",
      "args": ["path/to/agentic-stress-test/mcp/server.mjs"]
    }
  }
}

Runs that spend model calls are not offered to agents. They stay on the page and on the command line, behind the cost check. The site's own description for agents is at llms.txt.

Ask the assistant

This site is itself an agentic site. The assistant can explain the tester and what the checks mean. To run the tester's own checks, open the tester.

Connecting the assistant… If this message stays, the assistant could not load. The rest of the page still works.