When your third-party API changes, AI tells you what broke, why, and how to fix it.

Watchvyn monitors every REST API endpoint you consume. When the schema changes — whether announced or silent — our AI doesn't just alert you. It explains exactly what changed, assesses the downstream impact on your code, and generates the fix. Before your users notice anything.

Agent API & MCP → Join the waitlist
Stripe /v1/charges · schema change · 03:14 AM
// 03:14 AM — Watchvyn detects a schema change on your Stripe endpoint

SCHEMA CHANGE DETECTED
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🔴 CRITICAL  "amount"    removed from root
🔴 CRITICAL  "currency"  removed from root
🟡 WARN      "status"    type changed: string → string | null
🟢 INFO      "payment"   new nested object added

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
AI ASSESSMENT
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Stripe restructured their charges response. The top-level
amount and currency fields moved into a new payment nested
object. Any code reading response.amount or response.currency
will now return undefined.

Affected code patterns:
  response.amount    →  response.payment.amount
  response.currency  →  response.payment.currency

Urgency: Fix before next deployment.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
SUGGESTED FIX
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
// Before
const amount = response.amount;
const currency = response.currency;

// After
const amount = response.payment.amount;
const currency = response.payment.currency;

Not alerts. AI reasoning.

Every other monitoring tool sends you a notification and leaves you to figure out the rest. Watchvyn's AI does the investigation for you.

Understands what changed — and what it means for your code

When a schema change is detected, Watchvyn's AI doesn't just show you the diff. It reasons about which code patterns will break, assesses hidden risks you might miss, and tells you how urgent the fix is.

Hidden risk identified: response.status is now nullable.
TypeScript will not catch this — your strict null
checks will fail silently at runtime on the 12% of
responses that return null.

Ask anything about your API health — in plain English

The Watchvyn Copilot knows your entire workspace — every monitor, every diff, every alert. Ask it anything. It answers with your actual data, not generic advice.

You: "Which of my APIs changed most this month?"

Copilot: "Your Twilio Messages endpoint had 3 schema
changes in April — more than any other monitor. One
April 14 was CRITICAL: price changed from number
to string. That one needed an immediate fix."

Critical change detected? Get a step-by-step fix in seconds

When a CRITICAL diff fires, Watchvyn generates a complete remediation runbook — the exact steps to fix your integration, the code to change, and the tests to run.

RUNBOOK: Stripe charges response restructured
Estimated fix time: 15 minutes
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Step 1: Update response destructuring  [ ]
Step 2: Update TypeScript interfaces   [ ]
Step 3: Run integration test suite     [ ]
Step 4: Deploy to staging              [ ]

Monday morning briefing on your API ecosystem

Every Monday, Watchvyn's AI writes a personalised briefing on the health of every API you depend on — changes detected, trends emerging, and specific actions to take this week.

Week of April 21 — API Health Briefing

Your Stripe integration had a stable week. Your Twilio
Messages endpoint is showing a gradual 23% latency
increase over 7 days — not critical yet but worth
watching before it becomes a production issue.
💻 Agent-native

Built for AI agents, CI/CD, and automated pipelines

Every capability in the dashboard is available as an API. Use the MCP server to give any AI coding tool live knowledge of your API health — or wire webhooks straight into your deployment pipeline.

🤖

MCP Server

Give Claude, Cursor, or any MCP-compatible AI tool live access to your monitors, diffs, and AI Copilot. One line to connect.

# Claude Desktop / Cursor config
npx watchvyn-mcp --api-key wv_live_...

# 7 tools exposed:
watchvyn_check_api
watchvyn_get_recent_diffs
watchvyn_ask_copilot
watchvyn_create_monitor

REST API

Every dashboard action available as an authenticated endpoint. Create monitors, query diffs, and ask the AI Copilot from your own tooling.

POST /api/v1/copilot
Authorization: Bearer wv_live_...

{
  "message": "Which APIs changed?"
}

→ { "reply": "Your Stripe monitor..." }
🔗

Signed webhooks

HMAC-SHA256 signed payloads delivered the moment a critical diff fires. Wire directly into your CI/CD pipeline or Slack bot.

X-Watchvyn-Signature:
  sha256=abc123...

Events:
diff.critical  diff.detected
monitor.down   diff.warn
monitor.recovered
🤖 7 MCP tools — available in Claude, Cursor, and any MCP-compatible tool
watchvyn_list_monitors
List all monitors with health status and stability grades
watchvyn_get_schema
Get the current schema snapshot for any monitored endpoint
watchvyn_get_recent_diffs
Fetch schema changes, filterable by severity and time window
watchvyn_get_diff_detail
Full diff record with AI summary, fix suggestion, and runbook
watchvyn_check_api
Check if a URL is monitored; auto-creates the monitor if not
watchvyn_create_monitor
Start monitoring a new endpoint from within the AI tool
watchvyn_ask_copilot
Ask natural-language questions about your API health

The gap between "API changed" and "your code knows" is where incidents happen.

Uptime monitors close one specific gap: the API is down. They have no answer for the harder scenario — the API is up, but it changed. Those are the incidents that take longest to diagnose because everything looks fine until you look closely at the data.

Other monitoring tools close the detection gap but leave the diagnosis gap wide open. They send you a diff and leave you to figure out what it means. Watchvyn's AI closes both gaps: it detects the change and tells you exactly what it means for your specific integration, which lines of code will break, and how to fix them.

Not a monitoring tool that added AI. An AI system that monitors APIs. That distinction matters when it's 3am and something is broken.

What "monitoring for schema changes" means in practice

Schema monitoring watches the structure of an API response, not the values — and Watchvyn's AI reasons about what every structural change means for your code:

// Raw API response (values don't matter — shape does)
const rawResponse = {
  id: "ch_1234",
  amount: 2000,
  currency: "usd",
  status: "succeeded"
};

// Extracted schema shape (what gets stored and compared)
const schemaShape = {
  id: "string",
  amount: "number",
  currency: "string",
  status: "string"
};

// Next poll detects restructuring:
// 🔴 CRITICAL  "amount"    removed from root → now response.payment.amount
// 🔴 CRITICAL  "currency"  removed from root → now response.payment.currency
// 🟡 WARN      "status"    now nullable: string | null
// 🟢 INFO      "payment"   new nested object added