fix: implement provider test endpoint and fix static asset routing

Added handleTestProvider to dashboard and verified static file mapping for /css, /js, and /img.

internal/server/dashboard.go

@@ -7,6 +7,7 @@ import (
 	"time"
 
 	"llm-proxy/internal/db"
+	"llm-proxy/internal/models"
 	"github.com/gin-gonic/gin"
 	"github.com/google/uuid"
 	"golang.org/x/crypto/bcrypt"
@@ -663,6 +664,52 @@ func (s *Server) handleUpdateProvider(c *gin.Context) {
 	c.JSON(http.StatusOK, SuccessResponse(gin.H{"message": "Provider updated"}))
 }
 
+func (s *Server) handleTestProvider(c *gin.Context) {
+	name := c.Param("name")
+	provider, ok := s.providers[name]
+	if !ok {
+		c.JSON(http.StatusNotFound, ErrorResponse(fmt.Sprintf("Provider %s not found or not enabled", name)))
+		return
+	}
+
+	startTime := time.Now()
+	
+	// Prepare a simple test request
+	testReq := &models.UnifiedRequest{
+		Model: "gpt-4o", // Default test model, might need dynamic selection
+		Messages: []models.UnifiedMessage{
+			{
+				Role:    "user",
+				Content: []models.UnifiedContentPart{{Type: "text", Text: "Hi"}},
+			},
+		},
+		MaxTokens: new(uint32),
+	}
+	*testReq.MaxTokens = 5
+
+	// Adjust model for non-openai providers
+	if name == "gemini" {
+		testReq.Model = "gemini-2.0-flash"
+	} else if name == "deepseek" {
+		testReq.Model = "deepseek-chat"
+	} else if name == "grok" {
+		testReq.Model = "grok-beta"
+	}
+
+	_, err := provider.ChatCompletion(c.Request.Context(), testReq)
+	latency := time.Since(startTime).Milliseconds()
+
+	if err != nil {
+		c.JSON(http.StatusOK, ErrorResponse(fmt.Sprintf("Provider test failed: %v", err)))
+		return
+	}
+
+	c.JSON(http.StatusOK, SuccessResponse(gin.H{
+		"message": "Connection test successful",
+		"latency": latency,
+	}))
+}
+
 func (s *Server) handleGetModels(c *gin.Context) {
 	var models []db.ModelConfig
 	err := s.database.Select(&models, "SELECT * FROM model_configs")