Filters binary files in order to reduce API costs

Author: vinyas-bharadwaj

.gitignore

@@ -27,3 +27,4 @@ commit-helper.json
 HACKTOBERFEST_SETUP.md
 PR_DESCRIPTION.md
 commit
+test.txt

README.md

@@ -29,13 +29,14 @@ Looking to contribute? Check out:
 
 ✨ **AI-Powered Commit Messages** - Automatically generate meaningful commit messages  
 🔄 **Multiple LLM Support** - Choose between Google Gemini, Grok, Claude, ChatGPT, or Ollama (local)  
+🧪 **Dry Run Mode** - Preview prompts without making API calls  
 📝 **Context-Aware** - Analyzes staged and unstaged changes  
 📋 **Auto-Copy to Clipboard** - Generated messages are automatically copied for instant use  
 🎛️ **Interactive Review Flow** - Accept, regenerate with new styles, or open the message in your editor before committing  
 📊 **File Statistics Display** - Visual preview of changed files and line counts  
-🔒 **Smart Security Scrubbing** - Automatically removes API keys, passwords, and sensitive data from diffs  
+💡 **Smart Security Scrubbing** - Automatically removes API keys, passwords, and sensitive data from diffs  
 🚀 **Easy to Use** - Simple CLI interface with beautiful terminal UI  
-⚡ **Fast** - Quick generation of commit messages//img.shields.io/badge/Hacktoberfest-2025-orange.
+⚡️ **Fast** - Quick generation of commit messages
 
 ## Supported LLM Providers
 
@@ -112,6 +113,27 @@ Or if running from source:
 go run cmd/commit-msg/main.go .
 ```
 
+### Preview Mode (Dry Run)
+
+Preview what would be sent to the LLM without making an API call:
+
+```bash
+commit . --dry-run
+```
+
+This displays:
+- The LLM provider that would be used
+- The exact prompt that would be sent
+- File statistics and change summary
+- Estimated token count
+- All without consuming API credits or sharing data
+
+Perfect for:
+- 🐛 **Debugging** - See exactly what prompt is being sent
+- 💰 **Cost Control** - Review before consuming API credits
+- 🔒 **Privacy** - Verify what data would be shared with external APIs
+- 🧪 **Development** - Test prompt changes without API calls
+
 ### Setup LLM and API Key
 
 ```bash

cmd/cli/createMsg.go

@@ -26,7 +26,8 @@ import (
 
 // CreateCommitMsg launches the interactive flow for reviewing, regenerating,
 // editing, and accepting AI-generated commit messages in the current repo.
-func CreateCommitMsg() {
+// If dryRun is true, it displays the prompt without making an API call.
+func CreateCommitMsg(dryRun bool) {
 	// Validate COMMIT_LLM and required API keys
 	useLLM, err := store.DefaultLLMKey()
 	if err != nil {
@@ -63,8 +64,8 @@ func CreateCommitMsg() {
 	}
 
 	pterm.DefaultHeader.WithFullWidth().
-		WithBackgroundStyle(pterm.NewStyle(pterm.BgDarkGray)).
-		WithTextStyle(pterm.NewStyle(pterm.FgLightWhite)).
+		WithBackgroundStyle(pterm.NewStyle(pterm.BgCyan)).
+		WithTextStyle(pterm.NewStyle(pterm.FgBlack, pterm.Bold)).
 		Println("Commit Message Generator")
 
 	pterm.Println()
@@ -94,6 +95,13 @@ func CreateCommitMsg() {
 		return
 	}
 
+	// Handle dry-run mode: display what would be sent to LLM without making API call
+	if dryRun {
+		pterm.Println()
+		displayDryRunInfo(commitLLM, config, changes, apiKey)
+		return
+	}
+
 	pterm.Println()
 	spinnerGenerating, err := pterm.DefaultSpinner.
 		WithSequence("⠋", "⠙", "⠹", "⠸", "⠼", "⠴", "⠦", "⠧", "⠇", "⠏").
@@ -226,6 +234,22 @@ var (
 	}
 	errSelectionCancelled = errors.New("selection cancelled")
 )
+// resolveOllamaConfig returns the URL and model for Ollama, using environment variables as fallbacks
+func resolveOllamaConfig(apiKey string) (url, model string) {
+	url = apiKey
+	if strings.TrimSpace(url) == "" {
+		url = os.Getenv("OLLAMA_URL")
+		if url == "" {
+			url = "http://localhost:11434/api/generate"
+		}
+	}
+	model = os.Getenv("OLLAMA_MODEL")
+	if model == "" {
+		model = "llama3.1"
+	}
+	return url, model
+}
+
 
 func generateMessage(provider types.LLMProvider, config *types.Config, changes string, apiKey string, opts *types.GenerationOptions) (string, error) {
 	switch provider {
@@ -238,17 +262,7 @@ func generateMessage(provider types.LLMProvider, config *types.Config, changes s
 	case types.ProviderGroq:
 		return groq.GenerateCommitMessage(config, changes, apiKey, opts)
 	case types.ProviderOllama:
-		url := apiKey
-		if strings.TrimSpace(url) == "" {
-			url = os.Getenv("OLLAMA_URL")
-			if url == "" {
-				url = "http://localhost:11434/api/generate"
-			}
-		}
-		model := os.Getenv("OLLAMA_MODEL")
-		if model == "" {
-			model = "llama3.1"
-		}
+		url, model := resolveOllamaConfig(apiKey)
 		return ollama.GenerateCommitMessage(config, changes, url, model, opts)
 	default:
 		return grok.GenerateCommitMessage(config, changes, apiKey, opts)
@@ -425,3 +439,90 @@ func displayProviderError(provider types.LLMProvider, err error) {
 		pterm.Error.Printf("LLM API error: %v\n", err)
 	}
 }
+
+// displayDryRunInfo shows what would be sent to the LLM without making an API call
+func displayDryRunInfo(provider types.LLMProvider, config *types.Config, changes string, apiKey string) {
+	pterm.DefaultHeader.WithFullWidth().
+		WithBackgroundStyle(pterm.NewStyle(pterm.BgBlue)).
+		WithTextStyle(pterm.NewStyle(pterm.FgWhite, pterm.Bold)).
+		Println("DRY RUN MODE - Preview Only")
+
+	pterm.Println()
+	pterm.Info.Println("This is a dry-run. No API call will be made to the LLM provider.")
+	pterm.Println()
+
+	// Display provider information
+	pterm.DefaultSection.Println("LLM Provider Configuration")
+	providerInfo := [][]string{
+		{"Provider", provider.String()},
+	}
+
+	// Add provider-specific info
+	switch provider {
+	case types.ProviderOllama:
+		url, model := resolveOllamaConfig(apiKey)
+		providerInfo = append(providerInfo, []string{"Ollama URL", url})
+		providerInfo = append(providerInfo, []string{"Model", model})
+	case types.ProviderGrok:
+		providerInfo = append(providerInfo, []string{"API Endpoint", config.GrokAPI})
+		providerInfo = append(providerInfo, []string{"API Key", maskAPIKey(apiKey)})
+	default:
+		providerInfo = append(providerInfo, []string{"API Key", maskAPIKey(apiKey)})
+	}
+
+	pterm.DefaultTable.WithHasHeader(false).WithData(providerInfo).Render()
+
+	pterm.Println()
+
+	// Build and display the prompt
+	opts := &types.GenerationOptions{Attempt: 1}
+	prompt := types.BuildCommitPrompt(changes, opts)
+
+	pterm.DefaultSection.Println("Prompt That Would Be Sent")
+	pterm.Println()
+
+	// Display prompt in a box
+	promptBox := pterm.DefaultBox.
+		WithTitle("Full LLM Prompt").
+		WithTitleTopCenter().
+		WithBoxStyle(pterm.NewStyle(pterm.FgCyan))
+	promptBox.Println(prompt)
+
+	pterm.Println()
+
+	// Display changes statistics
+	pterm.DefaultSection.Println("Changes Summary")
+	linesCount := len(strings.Split(changes, "\n"))
+	charsCount := len(changes)
+
+	statsData := [][]string{
+		{"Total Lines", fmt.Sprintf("%d", linesCount)},
+		{"Total Characters", fmt.Sprintf("%d", charsCount)},
+		{"Prompt Size (approx)", fmt.Sprintf("%d tokens", estimateTokens(prompt))},
+	}
+	pterm.DefaultTable.WithHasHeader(false).WithData(statsData).Render()
+
+	pterm.Println()
+	pterm.Success.Println("Dry-run complete. To generate actual commit message, run without --dry-run flag.")
+}
+
+// maskAPIKey masks the API key for display purposes
+func maskAPIKey(apiKey string) string {
+	if len(apiKey) == 0 {
+		return "[NOT SET]"
+	}
+	// Don't mask URLs (used by Ollama)
+	if strings.HasPrefix(apiKey, "http://") || strings.HasPrefix(apiKey, "https://") {
+		return apiKey
+	}
+	if len(apiKey) <= 8 {
+		return strings.Repeat("*", len(apiKey))
+	}
+	// Show first 4 and last 4 characters
+	return apiKey[:4] + strings.Repeat("*", len(apiKey)-8) + apiKey[len(apiKey)-4:]
+}
+
+// estimateTokens provides a rough estimate of token count (1 token ≈ 4 characters)
+func estimateTokens(text string) int {
+	return len(text) / 4
+}

cmd/cli/root.go

@@ -53,7 +53,11 @@ var creatCommitMsg = &cobra.Command{
 	Use:   ".",
 	Short: "Create Commit Message",
 	RunE: func(cmd *cobra.Command, args []string) error {
-		CreateCommitMsg()
+		dryRun, err := cmd.Flags().GetBool("dry-run")
+		if err != nil {
+			return err
+		}
+		CreateCommitMsg(dryRun)
 		return nil
 	},
 }
@@ -68,6 +72,10 @@ func init() {
 	// Cobra also supports local flags, which will only run
 	// when this action is called directly.
 	rootCmd.Flags().BoolP("toggle", "t", false, "Help message for toggle")
+
+	// Add --dry-run flag to the commit command
+	creatCommitMsg.Flags().Bool("dry-run", false, "Preview the prompt that would be sent to the LLM without making an API call")
+
 	rootCmd.AddCommand(creatCommitMsg)
 	rootCmd.AddCommand(llmCmd)
 	llmCmd.AddCommand(llmSetupCmd)

go.mod

@@ -6,8 +6,8 @@ toolchain go1.24.7
 
 require (
 	github.com/atotto/clipboard v0.1.4
-	github.com/google/shlex v0.0.0-20191202100458-e7afc7fbc510
 	github.com/google/generative-ai-go v0.19.0
+	github.com/google/shlex v0.0.0-20191202100458-e7afc7fbc510
 	github.com/manifoldco/promptui v0.9.0
 	github.com/openai/openai-go/v3 v3.0.1
 	github.com/pterm/pterm v0.12.80

internal/display/display.go

@@ -112,10 +112,10 @@ func ShowCommitMessage(message string) {
 		WithBoxStyle(pterm.NewStyle(pterm.FgLightGreen)).
 		WithHorizontalString("─").
 		WithVerticalString("│").
-		WithTopLeftCornerString("┌").
-		WithTopRightCornerString("┐").
-		WithBottomLeftCornerString("└").
-		WithBottomRightCornerString("┘")
+		WithTopLeftCornerString("┘").
+		WithTopRightCornerString("└").
+		WithBottomLeftCornerString("┐").
+		WithBottomRightCornerString("┌")
 
 	panel.Println(pterm.LightGreen(message))
 }