@salimassili/ai-costguard 2.1.2 → 2.2.0

package/examples/integrations/webhook-alerts.mjs

@@ -0,0 +1,58 @@
+import { guard, GuardError, registerTokenizer } from '@salimassili/ai-costguard';
+
+const webhookUrl = process.env.COSTGUARD_WEBHOOK_URL;
+let providerCalls = 0;
+
+registerTokenizer('gpt-4o-mini', (text) => Math.ceil(text.length / 4));
+
+const mockedOpenAI = {
+  chat: {
+    completions: {
+      create: async () => {
+        providerCalls += 1;
+        return { ok: true };
+      },
+    },
+  },
+};
+
+const openai = guard(mockedOpenAI, {
+  budget: { maxUsd: 0.0001 },
+  projectId: 'webhook-alert-demo',
+  runId: 'local-run-1',
+  alerts: {
+    webhookUrl,
+    events: ['blocked'],
+    timeoutMs: 1000,
+  },
+});
+
+try {
+  await openai.chat.completions.create({
+    model: 'gpt-4o-mini',
+    messages: [{ role: 'user', content: 'Reserve a long output for an expensive agent step.' }],
+    max_tokens: 800,
+  });
+
+  throw new Error('Expected AI CostGuard to block the mocked provider call.');
+} catch (error) {
+  if (!(error instanceof GuardError)) throw error;
+
+  console.log(
+    JSON.stringify(
+      {
+        demo: 'webhook-alerts',
+        alert: webhookUrl ? 'enabled from COSTGUARD_WEBHOOK_URL' : 'off by default',
+        blocked: true,
+        reason: error.code,
+        providerCalls,
+      },
+      null,
+      2
+    )
+  );
+}
+
+if (providerCalls !== 0) {
+  throw new Error(`Provider executed ${providerCalls} time(s); expected 0.`);
+}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@salimassili/ai-costguard",
-  "version": "2.1.2",
+  "version": "2.2.0",
   "description": "Local-first runtime safety layer for AI agents that blocks runaway costs, loops, retries, and budget overruns before API calls execute.",
   "type": "module",
   "main": "./dist/index.js",