feat: add OpenAI-compat adapter for Ollama and llama.cpp

Thin wrapper over OpenAI adapter with custom base URLs.
Ollama: localhost:11434/v1, llama.cpp: localhost:8080/v1.
No API key required for local providers.

Fixed: initial tool call args captured on first chunk
(Ollama sends complete args in one chunk, not as deltas).

Live verified: text + tool calling with qwen3:14b on Ollama.
Five providers now live: Mistral, Anthropic, OpenAI, Google, Ollama.

cmd/gnoma/main.go

@@ -16,6 +16,7 @@ import (
 	"somegit.dev/Owlibou/gnoma/internal/provider/mistral"
 	googleprov "somegit.dev/Owlibou/gnoma/internal/provider/google"
 	oaiprov "somegit.dev/Owlibou/gnoma/internal/provider/openai"
+	"somegit.dev/Owlibou/gnoma/internal/provider/openaicompat"
 	"somegit.dev/Owlibou/gnoma/internal/stream"
 	"somegit.dev/Owlibou/gnoma/internal/tool"
 	"somegit.dev/Owlibou/gnoma/internal/tool/bash"
@@ -46,12 +47,13 @@ func main() {
 	}
 	logger := slog.New(slog.NewTextHandler(os.Stderr, &slog.HandlerOptions{Level: logLevel}))
 
-	// Resolve API key
+	// Resolve API key (local providers don't need one)
 	key := *apiKey
 	if key == "" {
 		key = resolveAPIKey(*providerName)
 	}
-	if key == "" {
+	localProviders := map[string]bool{"ollama": true, "llamacpp": true}
+	if key == "" && !localProviders[*providerName] {
 		fmt.Fprintf(os.Stderr, "error: no API key for provider %q\nSet %s environment variable or use --api-key\n",
 			*providerName, envKeyFor(*providerName))
 		os.Exit(1)
@@ -198,8 +200,12 @@ func createProvider(name, apiKey, model string) (provider.Provider, error) {
 		return oaiprov.New(cfg)
 	case "google":
 		return googleprov.New(cfg)
+	case "ollama":
+		return openaicompat.NewOllama(cfg)
+	case "llamacpp":
+		return openaicompat.NewLlamaCpp(cfg)
 	default:
-		return nil, fmt.Errorf("unknown provider %q (supports: mistral, anthropic, openai, google)", name)
+		return nil, fmt.Errorf("unknown provider %q (supports: mistral, anthropic, openai, google, ollama, llamacpp)", name)
 	}
 }
 

internal/provider/openai/stream.go

@@ -72,10 +72,11 @@ func (s *openaiStream) Next() bool {
 			for _, tc := range delta.ToolCalls {
 				existing, ok := s.toolCalls[tc.Index]
 				if !ok {
-					// New tool call
+					// New tool call — capture initial arguments too
 					existing = &toolCallState{
 						id:   tc.ID,
 						name: tc.Function.Name,
+						args: tc.Function.Arguments,
 					}
 					s.toolCalls[tc.Index] = existing
 					s.hadToolCalls = true
@@ -90,8 +91,8 @@ func (s *openaiStream) Next() bool {
 					}
 				}
 
-				// Accumulate arguments
-				if tc.Function.Arguments != "" {
+				// Accumulate arguments (subsequent chunks)
+				if tc.Function.Arguments != "" && ok {
 					existing.args += tc.Function.Arguments
 					s.cur = stream.Event{
 						Type:       stream.EventToolCallDelta,

internal/provider/openaicompat/provider.go

@@ -0,0 +1,41 @@
+// Package openaicompat provides OpenAI-compatible provider adapters
+// for Ollama, llama.cpp, and other servers that implement the OpenAI API.
+package openaicompat
+
+import (
+	"somegit.dev/Owlibou/gnoma/internal/provider"
+	oaiprov "somegit.dev/Owlibou/gnoma/internal/provider/openai"
+)
+
+const (
+	ollamaDefaultURL   = "http://localhost:11434/v1"
+	llamacppDefaultURL = "http://localhost:8080/v1"
+)
+
+// NewOllama creates a provider for a local Ollama instance.
+func NewOllama(cfg provider.ProviderConfig) (provider.Provider, error) {
+	if cfg.BaseURL == "" {
+		cfg.BaseURL = ollamaDefaultURL
+	}
+	if cfg.APIKey == "" {
+		cfg.APIKey = "ollama" // Ollama doesn't require a real key
+	}
+	if cfg.Model == "" {
+		cfg.Model = "qwen3:8b"
+	}
+	return oaiprov.New(cfg)
+}
+
+// NewLlamaCpp creates a provider for a local llama.cpp server.
+func NewLlamaCpp(cfg provider.ProviderConfig) (provider.Provider, error) {
+	if cfg.BaseURL == "" {
+		cfg.BaseURL = llamacppDefaultURL
+	}
+	if cfg.APIKey == "" {
+		cfg.APIKey = "llamacpp" // llama.cpp doesn't require a real key
+	}
+	if cfg.Model == "" {
+		cfg.Model = "default"
+	}
+	return oaiprov.New(cfg)
+}