chore(main): release 0.4.26 (#1584)

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>

.release-please-manifest.json

@@ -1 +1 @@
-{".":"0.4.24"}
+{".":"0.4.26"}

CHANGELOG.md

@@ -1,5 +1,19 @@
 # Changelog
 
+## [0.4.26](https://github.com/k8sgpt-ai/k8sgpt/compare/v0.4.25...v0.4.26) (2025-10-16)
+
+
+### Other
+
+* missing filter arg on serve ([#1583](https://github.com/k8sgpt-ai/k8sgpt/issues/1583)) ([f1d2e30](https://github.com/k8sgpt-ai/k8sgpt/commit/f1d2e306f32eb1e01a2788174084be29a7fa1282))
+
+## [0.4.25](https://github.com/k8sgpt-ai/k8sgpt/compare/v0.4.24...v0.4.25) (2025-09-03)
+
+
+### Features
+
+* fix to broken inference ([#1575](https://github.com/k8sgpt-ai/k8sgpt/issues/1575)) ([291e42d](https://github.com/k8sgpt-ai/k8sgpt/commit/291e42dc4b81ffb0672c21fbb325ddebc5d531a3))
+
 ## [0.4.24](https://github.com/k8sgpt-ai/k8sgpt/compare/v0.4.23...v0.4.24) (2025-08-18)
 
 

README.md

@@ -62,7 +62,7 @@ brew install k8sgpt
   <!---x-release-please-start-version-->
 
   ```
-  sudo rpm -ivh https://github.com/k8sgpt-ai/k8sgpt/releases/download/v0.4.24/k8sgpt_386.rpm
+  sudo rpm -ivh https://github.com/k8sgpt-ai/k8sgpt/releases/download/v0.4.26/k8sgpt_386.rpm
   ```
   <!---x-release-please-end-->
 
@@ -70,7 +70,7 @@ brew install k8sgpt
 
   <!---x-release-please-start-version-->
   ```
-  sudo rpm -ivh https://github.com/k8sgpt-ai/k8sgpt/releases/download/v0.4.24/k8sgpt_amd64.rpm
+  sudo rpm -ivh https://github.com/k8sgpt-ai/k8sgpt/releases/download/v0.4.26/k8sgpt_amd64.rpm
   ```
   <!---x-release-please-end-->
 </details>
@@ -83,7 +83,7 @@ brew install k8sgpt
   <!---x-release-please-start-version-->
 
 ```
-curl -LO https://github.com/k8sgpt-ai/k8sgpt/releases/download/v0.4.24/k8sgpt_386.deb
+curl -LO https://github.com/k8sgpt-ai/k8sgpt/releases/download/v0.4.26/k8sgpt_386.deb
 sudo dpkg -i k8sgpt_386.deb
 ```
 
@@ -94,7 +94,7 @@ sudo dpkg -i k8sgpt_386.deb
   <!---x-release-please-start-version-->
 
 ```
-curl -LO https://github.com/k8sgpt-ai/k8sgpt/releases/download/v0.4.24/k8sgpt_amd64.deb
+curl -LO https://github.com/k8sgpt-ai/k8sgpt/releases/download/v0.4.26/k8sgpt_amd64.deb
 sudo dpkg -i k8sgpt_amd64.deb
 ```
 
@@ -109,7 +109,7 @@ sudo dpkg -i k8sgpt_amd64.deb
 
   <!---x-release-please-start-version-->
   ```
-  wget https://github.com/k8sgpt-ai/k8sgpt/releases/download/v0.4.24/k8sgpt_386.apk
+  wget https://github.com/k8sgpt-ai/k8sgpt/releases/download/v0.4.26/k8sgpt_386.apk
   apk add --allow-untrusted k8sgpt_386.apk
   ```
   <!---x-release-please-end-->
@@ -118,7 +118,7 @@ sudo dpkg -i k8sgpt_amd64.deb
 
   <!---x-release-please-start-version-->
   ```
-  wget https://github.com/k8sgpt-ai/k8sgpt/releases/download/v0.4.24/k8sgpt_amd64.apk
+  wget https://github.com/k8sgpt-ai/k8sgpt/releases/download/v0.4.26/k8sgpt_amd64.apk
   apk add --allow-untrusted k8sgpt_amd64.apk
   ```
   <!---x-release-please-end-->

cmd/serve/serve.go

@@ -41,6 +41,8 @@ var (
 	enableMCP   bool
 	mcpPort     string
 	mcpHTTP     bool
+	// filters can be injected into the server (repeatable flag)
+	filters     []string
 )
 
 var ServeCmd = &cobra.Command{
@@ -208,6 +210,7 @@ var ServeCmd = &cobra.Command{
 			EnableHttp:  enableHttp,
 			Token:       aiProvider.Password,
 			Logger:      logger,
+			Filters:     filters,
 		}
 		go func() {
 			if err := server.ServeMetrics(); err != nil {
@@ -237,4 +240,6 @@ func init() {
 	ServeCmd.Flags().BoolVarP(&enableMCP, "mcp", "", false, "Enable Mission Control Protocol server")
 	ServeCmd.Flags().StringVarP(&mcpPort, "mcp-port", "", "8089", "Port to run the MCP server on")
 	ServeCmd.Flags().BoolVarP(&mcpHTTP, "mcp-http", "", false, "Enable HTTP mode for MCP server")
+	// allow injecting filters into the running server (repeatable)
+	ServeCmd.Flags().StringSliceVar(&filters, "filter", []string{}, "Filter to apply (can be specified multiple times)")
 }

pkg/ai/amazonbedrock.go

@@ -458,26 +458,25 @@ func (a *AmazonBedRockClient) Configure(config IAIConfig) error {
 		// Get the inference profile details
 		profile, err := a.getInferenceProfile(context.Background(), modelInput)
 		if err != nil {
-			// Instead of using a fallback model, throw an error
 			return fmt.Errorf("failed to get inference profile: %v", err)
-		} else {
-			// Extract the model ID from the inference profile
-			modelID, err := a.extractModelFromInferenceProfile(profile)
-			if err != nil {
-				return fmt.Errorf("failed to extract model ID from inference profile: %v", err)
-			}
-
-			// Find the model configuration for the extracted model ID
-			foundModel, err := a.getModelFromString(modelID)
-			if err != nil {
-				// Instead of using a fallback model, throw an error
-				return fmt.Errorf("failed to find model configuration for %s: %v", modelID, err)
-			}
-			a.model = foundModel
-
-			// Use the inference profile ARN as the model ID for API calls
-			a.model.Config.ModelName = modelInput
 		}
+		// Extract the model ID from the inference profile
+		modelID, err := a.extractModelFromInferenceProfile(profile)
+		if err != nil {
+			return fmt.Errorf("failed to extract model ID from inference profile: %v", err)
+		}
+		// Find the model configuration for the extracted model ID
+		foundModel, err := a.getModelFromString(modelID)
+		if err != nil {
+			// Instead of failing, use a generic config for completion/response
+			// But still warn user
+			return fmt.Errorf("failed to find model configuration for %s: %v", modelID, err)
+		}
+		// Use the found model config for completion/response, but set ModelName to the profile ARN
+		a.model = foundModel
+		a.model.Config.ModelName = modelInput
+		// Mark that we're using an inference profile
+		// (could add a field if needed)
 	} else {
 		// Regular model ID provided
 		foundModel, err := a.getModelFromString(modelInput)
@@ -562,7 +561,8 @@ func (a *AmazonBedRockClient) GetCompletion(ctx context.Context, prompt string)
 		supportedModels[i] = m.Name
 	}
 
-	if !bedrock_support.IsModelSupported(a.model.Config.ModelName, supportedModels) {
+	// Allow valid inference profile ARNs as supported models
+	if !bedrock_support.IsModelSupported(a.model.Config.ModelName, supportedModels) && !validateInferenceProfileArn(a.model.Config.ModelName) {
 		return "", fmt.Errorf("model '%s' is not supported.\nSupported models:\n%s", a.model.Config.ModelName, func() string {
 			s := ""
 			for _, m := range supportedModels {

pkg/server/server.go

@@ -54,6 +54,8 @@ type Config struct {
 	AnalyzeHandler *analyze.Handler
 	QueryHandler   *query.Handler
 	Logger         *zap.Logger
+	// Filters can be injected into the server to limit analysis to specific analyzers
+	Filters        []string
 	metricsServer  *http.Server
 	listener       net.Listener
 	EnableHttp     bool