feat: implement usage tracking for AI requests

Adds local token usage tracking for all AI providers. Usage data is
captured during stream processing and stored in a local database.
Includes a new Usage tab in the Providers dashboard to visualize
historical token consumption.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

src/sse/handlers/chat.js

@@ -120,6 +120,7 @@ async function handleSingleModelChat(body, modelStr, clientRawRequest = null) {
       credentials: refreshedCredentials,
       log,
       clientRawRequest,
+      connectionId: credentials.connectionId,
       onCredentialsRefreshed: async (newCreds) => {
         await updateProviderCredentials(credentials.connectionId, {
           accessToken: newCreds.accessToken,