Fix dead MCP server URLs in bundled plugins (apollo, outreach, datadog; remove servicenow) (#238)

- sales: apollo api.apollo.io/mcp -> mcp.apollo.io/mcp
- sales: outreach mcp.outreach.io/mcp -> api.outreach.io/mcp/
- engineering: datadog mcp.datadoghq.com/mcp -> mcp.datadoghq.com/api/unstable/mcp-server/mcp
- operations: remove servicenow (mcp.servicenow.com does not resolve;
  ServiceNow's MCP server is per-instance, no shared hosted URL exists)

All replacement URLs verified live (401/405 on bare GET, i.e. endpoint
exists and is auth-gated / POST-only). All replaced URLs verified dead
(404 or DNS NXDOMAIN). gcal/gmail intentionally untouched; see PR body.

engineering/.mcp.json

@@ -34,7 +34,7 @@
     },
     "datadog": {
       "type": "http",
-      "url": "https://mcp.datadoghq.com/mcp"
+      "url": "https://mcp.datadoghq.com/api/unstable/mcp-server/mcp"
     },
     "google calendar": {
       "type": "http",

operations/.mcp.json

@@ -28,10 +28,6 @@
       "type": "http",
       "url": "https://mcp.asana.com/v2/mcp"
     },
-    "servicenow": {
-      "type": "http",
-      "url": "https://mcp.servicenow.com/mcp"
-    },
     "ms365": {
       "type": "http",
       "url": "https://microsoft365.mcp.claude.com/mcp"

sales/.mcp.json

@@ -42,11 +42,11 @@
     },
     "apollo": {
       "type": "http",
-      "url": "https://api.apollo.io/mcp"
+      "url": "https://mcp.apollo.io/mcp"
     },
     "outreach": {
       "type": "http",
-      "url": "https://mcp.outreach.io/mcp"
+      "url": "https://api.outreach.io/mcp/"
     },
     "google calendar": {
       "type": "http",