checkpoint: provider compatibility

pkg/config/config.go

@@ -152,10 +152,12 @@ type ProvidersConfig struct {
 }
 
 type ProviderConfig struct {
-	APIKey     string `json:"api_key" env:"PICOCLAW_PROVIDERS_{{.Name}}_API_KEY"`
+	APIKey     string            `json:"api_key" env:"PICOCLAW_PROVIDERS_{{.Name}}_API_KEY"`
-	APIBase    string `json:"api_base" env:"PICOCLAW_PROVIDERS_{{.Name}}_API_BASE"`
+	APIBase    string            `json:"api_base" env:"PICOCLAW_PROVIDERS_{{.Name}}_API_BASE"`
-	Proxy      string `json:"proxy,omitempty" env:"PICOCLAW_PROVIDERS_{{.Name}}_PROXY"`
+	API        string            `json:"api,omitempty" env:"PICOCLAW_PROVIDERS_{{.Name}}_API"`
-	AuthMethod string `json:"auth_method,omitempty" env:"PICOCLAW_PROVIDERS_{{.Name}}_AUTH_METHOD"`
+	Headers    map[string]string `json:"headers,omitempty"`
+	Proxy      string            `json:"proxy,omitempty" env:"PICOCLAW_PROVIDERS_{{.Name}}_PROXY"`
+	AuthMethod string            `json:"auth_method,omitempty" env:"PICOCLAW_PROVIDERS_{{.Name}}_AUTH_METHOD"`
 }
 
 type GatewayConfig struct {

pkg/providers/http_provider.go

@@ -10,23 +10,39 @@ import (
 	"bytes"
 	"context"
 	"encoding/json"
+	"errors"
 	"fmt"
 	"io"
 	"net/http"
 	"net/url"
 	"strings"
 
+	"github.com/openai/openai-go/v3/responses"
 	"github.com/sipeed/picoclaw/pkg/auth"
 	"github.com/sipeed/picoclaw/pkg/config"
+	"github.com/sipeed/picoclaw/pkg/logger"
+	"github.com/sipeed/picoclaw/pkg/utils"
 )
 
 type HTTPProvider struct {
 	apiKey     string
 	apiBase    string
+	apiMode    string
+	headers    map[string]string
 	httpClient *http.Client
 }
 
-func NewHTTPProvider(apiKey, apiBase, proxy string) *HTTPProvider {
+type httpProviderError struct {
+	statusCode int
+	body       string
+	url        string
+}
+
+func (e *httpProviderError) Error() string {
+	return fmt.Sprintf("API error (%d): %s", e.statusCode, e.body)
+}
+
+func NewHTTPProvider(apiKey, apiBase, proxy, apiMode string, headers map[string]string) *HTTPProvider {
 	client := &http.Client{
 		Timeout: 0,
 	}
@@ -43,6 +59,8 @@ func NewHTTPProvider(apiKey, apiBase, proxy string) *HTTPProvider {
 	return &HTTPProvider{
 		apiKey:     apiKey,
 		apiBase:    apiBase,
+		apiMode:    apiMode,
+		headers:    headers,
 		httpClient: client,
 	}
 }
@@ -52,74 +70,32 @@ func (p *HTTPProvider) Chat(ctx context.Context, messages []Message, tools []Too
 		return nil, fmt.Errorf("API base not configured")
 	}
 
-	// Strip provider prefix from model name (e.g., moonshot/kimi-k2.5 -> kimi-k2.5)
+	useResponses := shouldPreferResponses(model, p.apiMode)
-	if idx := strings.Index(model, "/"); idx != -1 {
+	if useResponses {
-		prefix := model[:idx]
+		resp, err := p.chatWithResponses(ctx, messages, tools, model, options)
-		if prefix == "moonshot" || prefix == "nvidia" {
+		if err == nil {
-			model = model[idx+1:]
+			return resp, nil
 		}
-	}
+		if shouldFallbackFromResponses(err) {
-
+			logger.DebugCF("provider", "Responses endpoint unsupported, falling back to chat/completions", map[string]interface{}{
-	requestBody := map[string]interface{}{
+				"model": model,
-		"model":    model,
+			})
-		"messages": messages,
+			return p.chatWithCompletions(ctx, messages, tools, model, options)
-	}
-
-	if len(tools) > 0 {
-		requestBody["tools"] = tools
-		requestBody["tool_choice"] = "auto"
-	}
-
-	if maxTokens, ok := options["max_tokens"].(int); ok {
-		lowerModel := strings.ToLower(model)
-		if strings.Contains(lowerModel, "glm") || strings.Contains(lowerModel, "o1") {
-			requestBody["max_completion_tokens"] = maxTokens
-		} else {
-			requestBody["max_tokens"] = maxTokens
 		}
+		return nil, err
 	}
 
-	if temperature, ok := options["temperature"].(float64); ok {
+	resp, err := p.chatWithCompletions(ctx, messages, tools, model, options)
-		lowerModel := strings.ToLower(model)
+	if err == nil {
-		// Kimi k2 models only support temperature=1
+		return resp, nil
-		if strings.Contains(lowerModel, "kimi") && strings.Contains(lowerModel, "k2") {
-			requestBody["temperature"] = 1.0
-		} else {
-			requestBody["temperature"] = temperature
-		}
 	}
-
+	if shouldFallbackFromCompletions(err) {
-	jsonData, err := json.Marshal(requestBody)
+		logger.DebugCF("provider", "Chat/completions endpoint unsupported, falling back to responses", map[string]interface{}{
-	if err != nil {
+			"model": model,
-		return nil, fmt.Errorf("failed to marshal request: %w", err)
+		})
+		return p.chatWithResponses(ctx, messages, tools, model, options)
 	}
-
+	return nil, err
-	req, err := http.NewRequestWithContext(ctx, "POST", p.apiBase+"/chat/completions", bytes.NewReader(jsonData))
-	if err != nil {
-		return nil, fmt.Errorf("failed to create request: %w", err)
-	}
-
-	req.Header.Set("Content-Type", "application/json")
-	if p.apiKey != "" {
-		req.Header.Set("Authorization", "Bearer "+p.apiKey)
-	}
-
-	resp, err := p.httpClient.Do(req)
-	if err != nil {
-		return nil, fmt.Errorf("failed to send request: %w", err)
-	}
-	defer resp.Body.Close()
-
-	body, err := io.ReadAll(resp.Body)
-	if err != nil {
-		return nil, fmt.Errorf("failed to read response: %w", err)
-	}
-
-	if resp.StatusCode != http.StatusOK {
-		return nil, fmt.Errorf("API error: %s", string(body))
-	}
-
-	return p.parseResponse(body)
 }
 
 func (p *HTTPProvider) parseResponse(body []byte) (*LLMResponse, error) {
@@ -196,6 +172,202 @@ func (p *HTTPProvider) GetDefaultModel() string {
 	return ""
 }
 
+func (p *HTTPProvider) applyHeaders(req *http.Request) {
+	if len(p.headers) > 0 {
+		for k, v := range p.headers {
+			req.Header.Set(k, v)
+		}
+	}
+
+	if req.Header.Get("Content-Type") == "" {
+		req.Header.Set("Content-Type", "application/json")
+	}
+	if p.apiKey != "" && req.Header.Get("Authorization") == "" {
+		req.Header.Set("Authorization", "Bearer "+p.apiKey)
+	}
+}
+
+func (p *HTTPProvider) chatWithCompletions(ctx context.Context, messages []Message, tools []ToolDefinition, model string, options map[string]interface{}) (*LLMResponse, error) {
+	model = normalizeModelForHTTP(model)
+	requestBody := map[string]interface{}{
+		"model":    model,
+		"messages": messages,
+	}
+
+	if len(tools) > 0 {
+		requestBody["tools"] = tools
+		requestBody["tool_choice"] = "auto"
+	}
+
+	if maxTokens, ok := options["max_tokens"].(int); ok {
+		lowerModel := strings.ToLower(model)
+		if strings.Contains(lowerModel, "glm") || strings.Contains(lowerModel, "o1") {
+			requestBody["max_completion_tokens"] = maxTokens
+		} else {
+			requestBody["max_tokens"] = maxTokens
+		}
+	}
+
+	if temperature, ok := options["temperature"].(float64); ok {
+		lowerModel := strings.ToLower(model)
+		// Kimi k2 models only support temperature=1
+		if strings.Contains(lowerModel, "kimi") && strings.Contains(lowerModel, "k2") {
+			requestBody["temperature"] = 1.0
+		} else {
+			requestBody["temperature"] = temperature
+		}
+	}
+
+	jsonData, err := json.Marshal(requestBody)
+	if err != nil {
+		return nil, fmt.Errorf("failed to marshal request: %w", err)
+	}
+
+	req, err := http.NewRequestWithContext(ctx, "POST", p.apiBase+"/chat/completions", bytes.NewReader(jsonData))
+	if err != nil {
+		return nil, fmt.Errorf("failed to create request: %w", err)
+	}
+	logger.DebugCF("provider", "HTTP request", map[string]interface{}{
+		"url":    req.URL.String(),
+		"method": req.Method,
+	})
+
+	p.applyHeaders(req)
+
+	resp, err := p.httpClient.Do(req)
+	if err != nil {
+		return nil, fmt.Errorf("failed to send request: %w", err)
+	}
+	defer resp.Body.Close()
+
+	body, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return nil, fmt.Errorf("failed to read response: %w", err)
+	}
+
+	if resp.StatusCode != http.StatusOK {
+		logger.DebugCF("provider", "HTTP response error", map[string]interface{}{
+			"status": resp.StatusCode,
+			"body":   utils.Truncate(string(body), 500),
+		})
+		return nil, &httpProviderError{statusCode: resp.StatusCode, body: string(body), url: req.URL.String()}
+	}
+
+	return p.parseResponse(body)
+}
+
+func (p *HTTPProvider) chatWithResponses(ctx context.Context, messages []Message, tools []ToolDefinition, model string, options map[string]interface{}) (*LLMResponse, error) {
+	model = normalizeModelForHTTP(model)
+	params := buildCodexParams(messages, tools, model, stripTemperature(options))
+
+	jsonData, err := json.Marshal(params)
+	if err != nil {
+		return nil, fmt.Errorf("failed to marshal responses request: %w", err)
+	}
+
+	req, err := http.NewRequestWithContext(ctx, "POST", p.apiBase+"/responses", bytes.NewReader(jsonData))
+	if err != nil {
+		return nil, fmt.Errorf("failed to create request: %w", err)
+	}
+	logger.DebugCF("provider", "HTTP request", map[string]interface{}{
+		"url":    req.URL.String(),
+		"method": req.Method,
+	})
+
+	p.applyHeaders(req)
+
+	resp, err := p.httpClient.Do(req)
+	if err != nil {
+		return nil, fmt.Errorf("failed to send request: %w", err)
+	}
+	defer resp.Body.Close()
+
+	body, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return nil, fmt.Errorf("failed to read response: %w", err)
+	}
+
+	if resp.StatusCode != http.StatusOK {
+		logger.DebugCF("provider", "HTTP response error", map[string]interface{}{
+			"status": resp.StatusCode,
+			"body":   utils.Truncate(string(body), 500),
+		})
+		return nil, &httpProviderError{statusCode: resp.StatusCode, body: string(body), url: req.URL.String()}
+	}
+
+	var apiResponse responses.Response
+	if err := json.Unmarshal(body, &apiResponse); err != nil {
+		return nil, fmt.Errorf("failed to unmarshal responses API response: %w", err)
+	}
+	return parseCodexResponse(&apiResponse), nil
+}
+
+func normalizeModelForHTTP(model string) string {
+	// Strip provider prefix from model name (e.g., moonshot/kimi-k2.5 -> kimi-k2.5)
+	if idx := strings.Index(model, "/"); idx != -1 {
+		prefix := model[:idx]
+		if prefix == "moonshot" || prefix == "nvidia" {
+			return model[idx+1:]
+		}
+	}
+	return model
+}
+
+func stripTemperature(options map[string]interface{}) map[string]interface{} {
+	if options == nil {
+		return nil
+	}
+	if _, ok := options["temperature"]; !ok {
+		return options
+	}
+	cleaned := make(map[string]interface{}, len(options)-1)
+	for k, v := range options {
+		if k == "temperature" {
+			continue
+		}
+		cleaned[k] = v
+	}
+	return cleaned
+}
+
+func shouldPreferResponses(model, apiMode string) bool {
+	lowerMode := strings.ToLower(apiMode)
+	switch lowerMode {
+	case "openai-responses", "responses", "response":
+		return true
+	case "openai-completions", "chat-completions", "completions":
+		return false
+	}
+
+	lower := strings.ToLower(model)
+	return strings.Contains(lower, "gpt-5") || strings.Contains(lower, "codex") || strings.Contains(lower, "o1")
+}
+
+func shouldFallbackFromResponses(err error) bool {
+	var httpErr *httpProviderError
+	if errors.As(err, &httpErr) {
+		return isEndpointUnsupported(httpErr.statusCode)
+	}
+	return false
+}
+
+func shouldFallbackFromCompletions(err error) bool {
+	var httpErr *httpProviderError
+	if errors.As(err, &httpErr) {
+		return isEndpointUnsupported(httpErr.statusCode)
+	}
+	return false
+}
+
+func isEndpointUnsupported(statusCode int) bool {
+	switch statusCode {
+	case http.StatusNotFound, http.StatusMethodNotAllowed, http.StatusNotImplemented, http.StatusGone:
+		return true
+	default:
+		return false
+	}
+}
+
 func createClaudeAuthProvider() (LLMProvider, error) {
 	cred, err := auth.GetCredential("anthropic")
 	if err != nil {
@@ -222,7 +394,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 	model := cfg.Agents.Defaults.Model
 	providerName := strings.ToLower(cfg.Agents.Defaults.Provider)
 
-	var apiKey, apiBase, proxy string
+	var apiKey, apiBase, proxy, apiMode string
+	var headers map[string]string
 
 	lowerModel := strings.ToLower(model)
 
@@ -233,6 +406,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 			if cfg.Providers.Groq.APIKey != "" {
 				apiKey = cfg.Providers.Groq.APIKey
 				apiBase = cfg.Providers.Groq.APIBase
+				apiMode = cfg.Providers.Groq.API
+				headers = cfg.Providers.Groq.Headers
 				if apiBase == "" {
 					apiBase = "https://api.groq.com/openai/v1"
 				}
@@ -244,6 +419,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 				}
 				apiKey = cfg.Providers.OpenAI.APIKey
 				apiBase = cfg.Providers.OpenAI.APIBase
+				apiMode = cfg.Providers.OpenAI.API
+				headers = cfg.Providers.OpenAI.Headers
 				if apiBase == "" {
 					apiBase = "https://api.openai.com/v1"
 				}
@@ -255,6 +432,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 				}
 				apiKey = cfg.Providers.Anthropic.APIKey
 				apiBase = cfg.Providers.Anthropic.APIBase
+				apiMode = cfg.Providers.Anthropic.API
+				headers = cfg.Providers.Anthropic.Headers
 				if apiBase == "" {
 					apiBase = "https://api.anthropic.com/v1"
 				}
@@ -262,6 +441,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 		case "openrouter":
 			if cfg.Providers.OpenRouter.APIKey != "" {
 				apiKey = cfg.Providers.OpenRouter.APIKey
+				apiMode = cfg.Providers.OpenRouter.API
+				headers = cfg.Providers.OpenRouter.Headers
 				if cfg.Providers.OpenRouter.APIBase != "" {
 					apiBase = cfg.Providers.OpenRouter.APIBase
 				} else {
@@ -272,6 +453,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 			if cfg.Providers.Zhipu.APIKey != "" {
 				apiKey = cfg.Providers.Zhipu.APIKey
 				apiBase = cfg.Providers.Zhipu.APIBase
+				apiMode = cfg.Providers.Zhipu.API
+				headers = cfg.Providers.Zhipu.Headers
 				if apiBase == "" {
 					apiBase = "https://open.bigmodel.cn/api/paas/v4"
 				}
@@ -280,6 +463,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 			if cfg.Providers.Gemini.APIKey != "" {
 				apiKey = cfg.Providers.Gemini.APIKey
 				apiBase = cfg.Providers.Gemini.APIBase
+				apiMode = cfg.Providers.Gemini.API
+				headers = cfg.Providers.Gemini.Headers
 				if apiBase == "" {
 					apiBase = "https://generativelanguage.googleapis.com/v1beta"
 				}
@@ -288,6 +473,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 			if cfg.Providers.VLLM.APIBase != "" {
 				apiKey = cfg.Providers.VLLM.APIKey
 				apiBase = cfg.Providers.VLLM.APIBase
+				apiMode = cfg.Providers.VLLM.API
+				headers = cfg.Providers.VLLM.Headers
 			}
 		case "claude-cli", "claudecode", "claude-code":
 			workspace := cfg.Agents.Defaults.Workspace
@@ -305,6 +492,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 			apiKey = cfg.Providers.Moonshot.APIKey
 			apiBase = cfg.Providers.Moonshot.APIBase
 			proxy = cfg.Providers.Moonshot.Proxy
+			apiMode = cfg.Providers.Moonshot.API
+			headers = cfg.Providers.Moonshot.Headers
 			if apiBase == "" {
 				apiBase = "https://api.moonshot.cn/v1"
 			}
@@ -312,6 +501,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 		case strings.HasPrefix(model, "openrouter/") || strings.HasPrefix(model, "anthropic/") || strings.HasPrefix(model, "openai/") || strings.HasPrefix(model, "meta-llama/") || strings.HasPrefix(model, "deepseek/") || strings.HasPrefix(model, "google/"):
 			apiKey = cfg.Providers.OpenRouter.APIKey
 			proxy = cfg.Providers.OpenRouter.Proxy
+			apiMode = cfg.Providers.OpenRouter.API
+			headers = cfg.Providers.OpenRouter.Headers
 			if cfg.Providers.OpenRouter.APIBase != "" {
 				apiBase = cfg.Providers.OpenRouter.APIBase
 			} else {
@@ -325,6 +516,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 			apiKey = cfg.Providers.Anthropic.APIKey
 			apiBase = cfg.Providers.Anthropic.APIBase
 			proxy = cfg.Providers.Anthropic.Proxy
+			apiMode = cfg.Providers.Anthropic.API
+			headers = cfg.Providers.Anthropic.Headers
 			if apiBase == "" {
 				apiBase = "https://api.anthropic.com/v1"
 			}
@@ -336,6 +529,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 			apiKey = cfg.Providers.OpenAI.APIKey
 			apiBase = cfg.Providers.OpenAI.APIBase
 			proxy = cfg.Providers.OpenAI.Proxy
+			apiMode = cfg.Providers.OpenAI.API
+			headers = cfg.Providers.OpenAI.Headers
 			if apiBase == "" {
 				apiBase = "https://api.openai.com/v1"
 			}
@@ -344,6 +539,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 			apiKey = cfg.Providers.Gemini.APIKey
 			apiBase = cfg.Providers.Gemini.APIBase
 			proxy = cfg.Providers.Gemini.Proxy
+			apiMode = cfg.Providers.Gemini.API
+			headers = cfg.Providers.Gemini.Headers
 			if apiBase == "" {
 				apiBase = "https://generativelanguage.googleapis.com/v1beta"
 			}
@@ -352,6 +549,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 			apiKey = cfg.Providers.Zhipu.APIKey
 			apiBase = cfg.Providers.Zhipu.APIBase
 			proxy = cfg.Providers.Zhipu.Proxy
+			apiMode = cfg.Providers.Zhipu.API
+			headers = cfg.Providers.Zhipu.Headers
 			if apiBase == "" {
 				apiBase = "https://open.bigmodel.cn/api/paas/v4"
 			}
@@ -360,6 +559,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 			apiKey = cfg.Providers.Groq.APIKey
 			apiBase = cfg.Providers.Groq.APIBase
 			proxy = cfg.Providers.Groq.Proxy
+			apiMode = cfg.Providers.Groq.API
+			headers = cfg.Providers.Groq.Headers
 			if apiBase == "" {
 				apiBase = "https://api.groq.com/openai/v1"
 			}
@@ -368,6 +569,8 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 			apiKey = cfg.Providers.Nvidia.APIKey
 			apiBase = cfg.Providers.Nvidia.APIBase
 			proxy = cfg.Providers.Nvidia.Proxy
+			apiMode = cfg.Providers.Nvidia.API
+			headers = cfg.Providers.Nvidia.Headers
 			if apiBase == "" {
 				apiBase = "https://integrate.api.nvidia.com/v1"
 			}
@@ -376,11 +579,15 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 			apiKey = cfg.Providers.VLLM.APIKey
 			apiBase = cfg.Providers.VLLM.APIBase
 			proxy = cfg.Providers.VLLM.Proxy
+			apiMode = cfg.Providers.VLLM.API
+			headers = cfg.Providers.VLLM.Headers
 
 		default:
 			if cfg.Providers.OpenRouter.APIKey != "" {
 				apiKey = cfg.Providers.OpenRouter.APIKey
 				proxy = cfg.Providers.OpenRouter.Proxy
+				apiMode = cfg.Providers.OpenRouter.API
+				headers = cfg.Providers.OpenRouter.Headers
 				if cfg.Providers.OpenRouter.APIBase != "" {
 					apiBase = cfg.Providers.OpenRouter.APIBase
 				} else {
@@ -400,5 +607,5 @@ func CreateProvider(cfg *config.Config) (LLMProvider, error) {
 		return nil, fmt.Errorf("no API base configured for provider (model: %s)", model)
 	}
 
-	return NewHTTPProvider(apiKey, apiBase, proxy), nil
+	return NewHTTPProvider(apiKey, apiBase, proxy, apiMode, headers), nil
 }