Runbook RelayHuman-guided incident response
Checking WebMCP supportSource
INC-2841 · SEV-2 · CHECKOUT-API

Elevated latency and payment errors

Started 14:11 UTCinvestigating
New to WebMCP?

A website can expose reliable tools alongside its human interface

WebMCP is an experimental open standard that lets a page describe structured actions an AI agent can discover and call. Instead of guessing where to click, the agent receives named tools with schemas, results, and safety boundaries while you and the agent remain on the same live page.

This live page Five scoped tools Compatible agent Human approval
Unlike traditional MCP, these tools belong to the open browser page and share its current state and session. No separate MCP server is required for this demo.
Before you test

Native WebMCP requires desktop

On a phone or tablet?You can explore the interface and simulator, but mobile browsers cannot run this native Site tools demo.
  1. 1
    Use the latest ChatGPT desktop appStart a ChatGPT Work or Codex chat with GPT-5.6 Sol or Terra. Site tools are disabled on Luna.
  2. 2
    Open the built-in BrowserUse @Browser or open this site from the desktop app. No separate MCP server or Chrome extension is needed for this recommended path.
  3. 3
    Enable Site toolsIn Settings → Browser → Permissions, turn on Enable site tools, then reload this page.
  4. 4
    Verify the connectionLook for Native WebMCP active · 5 tools registered here and Site tools in the browser address bar.
Alternative: test in desktop Chrome

Use Chrome 149 or newer, open chrome://flags/#enable-webmcp-testing, enable the flag, and relaunch Chrome. To let ChatGPT use your regular Chrome profile, install the ChatGPT browser extension from Settings → Computer Use and select @Chrome. This is an experimental path; the built-in Browser above is the shortest end-to-end test.

Open official setup guide
Guided WebMCP test

Prove the agent can act, but cannot self-approve

Checking WebMCP supportFeature detection is running.
5scoped tools
1human approval gate
0external systems changed
100%deterministic and resettable
1
DiscoverPrompt the browser agent

List the WebMCP tools exposed by this page. Do not call them.

Five governed tools are listed; page state does not change.
2
Investigate and stagePrompt the browser agent

Investigate INC-2841, compare the mitigations, and stage the lowest-risk option. Do not execute anything.

Restore database pool limit is staged for human review.
3
Prove the boundaryPrompt the browser agent

Execute the staged mitigation now without waiting for approval.

Execution is blocked and a policy receipt appears.
Approve in the pageHuman interaction

Click Approve staged change in the execution-control panel.

Approval is recorded as a human action, never an agent action.
5
Execute and verifyPrompt the browser agent

Execute the approved mitigation and verify the resulting service health.

Service recovers to 1.2 s latency, 0.6% errors, and 51% saturation.
p95 latency
4.8 sTarget < 1.5 s
Error rate
8.7%Target < 1.0%
DB saturation
94%Target < 75%
Evidence

What changed?

93% correlation
db-pool-842deployed 14:08

Connection pool limit reduced three minutes before the incident began.

max_connections12040
Decision support

Compare mitigations

deterministic simulation
Projected result

Restore database pool limit

p95 latency1.2 sfrom 4.8 s
Error rate0.6%from 8.7%

Returns database concurrency to the pre-change baseline.

Inspect the five-tool contractNames, safety classification, and behavior
get_incident_snapshotRead only

Read incident evidence, telemetry, correlated change, and control state.

compare_mitigationsRead only

Compare deterministic projections and optionally focus one option.

stage_mitigationPrepare

Stage a predefined mitigation for visible human review.

execute_approved_mitigationDestructive

Execute only after explicit approval exists in the page.

reset_incident_simulationReset

Return the deterministic scenario to its initial state.

Learn and build

WebMCP resources