Merge pull request #1428 from YanzheL/fix/openai-gateway-content-session-hash-fallback

fix(gateway): add content-based session hash fallback for non-Codex clients

backend/internal/service/openai_content_session_seed.go

@@ -0,0 +1,107 @@
+package service
+
+import (
+	"encoding/json"
+	"strings"
+
+	"github.com/tidwall/gjson"
+)
+
+// contentSessionSeedPrefix prevents collisions between content-derived seeds
+// and explicit session IDs (e.g. "sess-xxx" or "compat_cc_xxx").
+const contentSessionSeedPrefix = "compat_cs_"
+
+// deriveOpenAIContentSessionSeed builds a stable session seed from an
+// OpenAI-format request body. Only fields constant across conversation turns
+// are included: model, tools/functions definitions, system/developer prompts,
+// instructions (Responses API), and the first user message.
+// Supports both Chat Completions (messages) and Responses API (input).
+func deriveOpenAIContentSessionSeed(body []byte) string {
+	if len(body) == 0 {
+		return ""
+	}
+
+	var b strings.Builder
+
+	if model := gjson.GetBytes(body, "model").String(); model != "" {
+		_, _ = b.WriteString("model=")
+		_, _ = b.WriteString(model)
+	}
+
+	if tools := gjson.GetBytes(body, "tools"); tools.Exists() && tools.IsArray() && tools.Raw != "[]" {
+		_, _ = b.WriteString("|tools=")
+		_, _ = b.WriteString(normalizeCompatSeedJSON(json.RawMessage(tools.Raw)))
+	}
+
+	if funcs := gjson.GetBytes(body, "functions"); funcs.Exists() && funcs.IsArray() && funcs.Raw != "[]" {
+		_, _ = b.WriteString("|functions=")
+		_, _ = b.WriteString(normalizeCompatSeedJSON(json.RawMessage(funcs.Raw)))
+	}
+
+	if instr := gjson.GetBytes(body, "instructions").String(); instr != "" {
+		_, _ = b.WriteString("|instructions=")
+		_, _ = b.WriteString(instr)
+	}
+
+	firstUserCaptured := false
+
+	msgs := gjson.GetBytes(body, "messages")
+	if msgs.Exists() && msgs.IsArray() {
+		msgs.ForEach(func(_, msg gjson.Result) bool {
+			role := msg.Get("role").String()
+			switch role {
+			case "system", "developer":
+				_, _ = b.WriteString("|system=")
+				if c := msg.Get("content"); c.Exists() {
+					_, _ = b.WriteString(normalizeCompatSeedJSON(json.RawMessage(c.Raw)))
+				}
+			case "user":
+				if !firstUserCaptured {
+					_, _ = b.WriteString("|first_user=")
+					if c := msg.Get("content"); c.Exists() {
+						_, _ = b.WriteString(normalizeCompatSeedJSON(json.RawMessage(c.Raw)))
+					}
+					firstUserCaptured = true
+				}
+			}
+			return true
+		})
+	} else if inp := gjson.GetBytes(body, "input"); inp.Exists() {
+		if inp.Type == gjson.String {
+			_, _ = b.WriteString("|input=")
+			_, _ = b.WriteString(inp.String())
+		} else if inp.IsArray() {
+			inp.ForEach(func(_, item gjson.Result) bool {
+				role := item.Get("role").String()
+				switch role {
+				case "system", "developer":
+					_, _ = b.WriteString("|system=")
+					if c := item.Get("content"); c.Exists() {
+						_, _ = b.WriteString(normalizeCompatSeedJSON(json.RawMessage(c.Raw)))
+					}
+				case "user":
+					if !firstUserCaptured {
+						_, _ = b.WriteString("|first_user=")
+						if c := item.Get("content"); c.Exists() {
+							_, _ = b.WriteString(normalizeCompatSeedJSON(json.RawMessage(c.Raw)))
+						}
+						firstUserCaptured = true
+					}
+				}
+				if !firstUserCaptured && item.Get("type").String() == "input_text" {
+					_, _ = b.WriteString("|first_user=")
+					if text := item.Get("text").String(); text != "" {
+						_, _ = b.WriteString(text)
+					}
+					firstUserCaptured = true
+				}
+				return true
+			})
+		}
+	}
+
+	if b.Len() == 0 {
+		return ""
+	}
+	return contentSessionSeedPrefix + b.String()
+}

backend/internal/service/openai_content_session_seed_test.go

@@ -0,0 +1,218 @@
+package service
+
+import (
+	"testing"
+
+	"github.com/stretchr/testify/require"
+)
+
+func TestDeriveOpenAIContentSessionSeed_EmptyInputs(t *testing.T) {
+	require.Empty(t, deriveOpenAIContentSessionSeed(nil))
+	require.Empty(t, deriveOpenAIContentSessionSeed([]byte{}))
+	require.Empty(t, deriveOpenAIContentSessionSeed([]byte(`{}`)))
+}
+
+func TestDeriveOpenAIContentSessionSeed_ModelOnly(t *testing.T) {
+	seed := deriveOpenAIContentSessionSeed([]byte(`{"model":"gpt-5.4"}`))
+	require.Contains(t, seed, contentSessionSeedPrefix)
+	require.Contains(t, seed, "model=gpt-5.4")
+}
+
+func TestDeriveOpenAIContentSessionSeed_ChatCompletions_StableAcrossTurns(t *testing.T) {
+	turn1 := []byte(`{
+		"model": "gpt-5.4",
+		"messages": [
+			{"role": "system", "content": "You are helpful."},
+			{"role": "user", "content": "Hello"}
+		]
+	}`)
+	turn2 := []byte(`{
+		"model": "gpt-5.4",
+		"messages": [
+			{"role": "system", "content": "You are helpful."},
+			{"role": "user", "content": "Hello"},
+			{"role": "assistant", "content": "Hi there!"},
+			{"role": "user", "content": "How are you?"}
+		]
+	}`)
+	s1 := deriveOpenAIContentSessionSeed(turn1)
+	s2 := deriveOpenAIContentSessionSeed(turn2)
+	require.Equal(t, s1, s2, "seed should be stable across later turns")
+	require.NotEmpty(t, s1)
+}
+
+func TestDeriveOpenAIContentSessionSeed_ChatCompletions_DifferentFirstUserDiffers(t *testing.T) {
+	req1 := []byte(`{"model":"gpt-5.4","messages":[{"role":"user","content":"Question A"}]}`)
+	req2 := []byte(`{"model":"gpt-5.4","messages":[{"role":"user","content":"Question B"}]}`)
+	s1 := deriveOpenAIContentSessionSeed(req1)
+	s2 := deriveOpenAIContentSessionSeed(req2)
+	require.NotEqual(t, s1, s2)
+}
+
+func TestDeriveOpenAIContentSessionSeed_ChatCompletions_DifferentSystemDiffers(t *testing.T) {
+	req1 := []byte(`{"model":"gpt-5.4","messages":[{"role":"system","content":"A"},{"role":"user","content":"Hi"}]}`)
+	req2 := []byte(`{"model":"gpt-5.4","messages":[{"role":"system","content":"B"},{"role":"user","content":"Hi"}]}`)
+	s1 := deriveOpenAIContentSessionSeed(req1)
+	s2 := deriveOpenAIContentSessionSeed(req2)
+	require.NotEqual(t, s1, s2)
+}
+
+func TestDeriveOpenAIContentSessionSeed_ChatCompletions_DifferentModelDiffers(t *testing.T) {
+	req1 := []byte(`{"model":"gpt-5.4","messages":[{"role":"user","content":"Hi"}]}`)
+	req2 := []byte(`{"model":"gpt-4o","messages":[{"role":"user","content":"Hi"}]}`)
+	s1 := deriveOpenAIContentSessionSeed(req1)
+	s2 := deriveOpenAIContentSessionSeed(req2)
+	require.NotEqual(t, s1, s2)
+}
+
+func TestDeriveOpenAIContentSessionSeed_ChatCompletions_WithTools(t *testing.T) {
+	withTools := []byte(`{
+		"model": "gpt-5.4",
+		"tools": [{"type":"function","function":{"name":"get_weather"}}],
+		"messages": [{"role": "user", "content": "Hello"}]
+	}`)
+	withoutTools := []byte(`{
+		"model": "gpt-5.4",
+		"messages": [{"role": "user", "content": "Hello"}]
+	}`)
+	s1 := deriveOpenAIContentSessionSeed(withTools)
+	s2 := deriveOpenAIContentSessionSeed(withoutTools)
+	require.NotEqual(t, s1, s2, "tools should affect the seed")
+	require.Contains(t, s1, "|tools=")
+}
+
+func TestDeriveOpenAIContentSessionSeed_ChatCompletions_WithFunctions(t *testing.T) {
+	body := []byte(`{
+		"model": "gpt-5.4",
+		"functions": [{"name":"get_weather","parameters":{}}],
+		"messages": [{"role": "user", "content": "Hello"}]
+	}`)
+	seed := deriveOpenAIContentSessionSeed(body)
+	require.Contains(t, seed, "|functions=")
+}
+
+func TestDeriveOpenAIContentSessionSeed_ChatCompletions_DeveloperRole(t *testing.T) {
+	body := []byte(`{
+		"model": "gpt-5.4",
+		"messages": [
+			{"role": "developer", "content": "You are helpful."},
+			{"role": "user", "content": "Hello"}
+		]
+	}`)
+	seed := deriveOpenAIContentSessionSeed(body)
+	require.Contains(t, seed, "|system=")
+	require.Contains(t, seed, "|first_user=")
+}
+
+func TestDeriveOpenAIContentSessionSeed_ChatCompletions_StructuredContent(t *testing.T) {
+	body := []byte(`{
+		"model": "gpt-5.4",
+		"messages": [
+			{"role": "user", "content": [{"type":"text","text":"Hello"}]}
+		]
+	}`)
+	seed := deriveOpenAIContentSessionSeed(body)
+	require.NotEmpty(t, seed)
+	require.Contains(t, seed, "|first_user=")
+}
+
+func TestDeriveOpenAIContentSessionSeed_ResponsesAPI_InputString(t *testing.T) {
+	body := []byte(`{"model":"gpt-5.4","input":"Hello, how are you?"}`)
+	seed := deriveOpenAIContentSessionSeed(body)
+	require.Contains(t, seed, "|input=Hello, how are you?")
+}
+
+func TestDeriveOpenAIContentSessionSeed_ResponsesAPI_InputArray(t *testing.T) {
+	body := []byte(`{
+		"model": "gpt-5.4",
+		"input": [
+			{"role": "system", "content": "You are helpful."},
+			{"role": "user", "content": "Hello"}
+		]
+	}`)
+	seed := deriveOpenAIContentSessionSeed(body)
+	require.Contains(t, seed, "|system=")
+	require.Contains(t, seed, "|first_user=")
+}
+
+func TestDeriveOpenAIContentSessionSeed_ResponsesAPI_WithInstructions(t *testing.T) {
+	body := []byte(`{
+		"model": "gpt-5.4",
+		"instructions": "You are a coding assistant.",
+		"input": "Write a hello world"
+	}`)
+	seed := deriveOpenAIContentSessionSeed(body)
+	require.Contains(t, seed, "|instructions=You are a coding assistant.")
+	require.Contains(t, seed, "|input=Write a hello world")
+}
+
+func TestDeriveOpenAIContentSessionSeed_Deterministic(t *testing.T) {
+	body := []byte(`{
+		"model": "gpt-5.4",
+		"messages": [
+			{"role": "system", "content": "You are helpful."},
+			{"role": "user", "content": "Hello"}
+		]
+	}`)
+	s1 := deriveOpenAIContentSessionSeed(body)
+	s2 := deriveOpenAIContentSessionSeed(body)
+	require.Equal(t, s1, s2, "seed must be deterministic")
+}
+
+func TestDeriveOpenAIContentSessionSeed_PrefixPresent(t *testing.T) {
+	body := []byte(`{"model":"gpt-5.4","messages":[{"role":"user","content":"Hi"}]}`)
+	seed := deriveOpenAIContentSessionSeed(body)
+	require.True(t, len(seed) > len(contentSessionSeedPrefix))
+	require.Equal(t, contentSessionSeedPrefix, seed[:len(contentSessionSeedPrefix)])
+}
+
+func TestDeriveOpenAIContentSessionSeed_EmptyToolsIgnored(t *testing.T) {
+	body := []byte(`{"model":"gpt-5.4","tools":[],"messages":[{"role":"user","content":"Hi"}]}`)
+	seed := deriveOpenAIContentSessionSeed(body)
+	require.NotContains(t, seed, "|tools=")
+}
+
+func TestDeriveOpenAIContentSessionSeed_MessagesPreferredOverInput(t *testing.T) {
+	body := []byte(`{
+		"model": "gpt-5.4",
+		"messages": [{"role": "user", "content": "from messages"}],
+		"input": "from input"
+	}`)
+	seed := deriveOpenAIContentSessionSeed(body)
+	require.Contains(t, seed, "|first_user=")
+	require.NotContains(t, seed, "|input=")
+}
+
+func TestDeriveOpenAIContentSessionSeed_JSONCanonicalisation(t *testing.T) {
+	compact := []byte(`{"model":"gpt-5.4","tools":[{"type":"function","function":{"name":"get_weather","description":"Get weather"}}],"messages":[{"role":"user","content":"Hi"}]}`)
+	spaced := []byte(`{
+		"model": "gpt-5.4",
+		"tools": [
+			{ "type" : "function", "function": { "description": "Get weather", "name": "get_weather" } }
+		],
+		"messages": [ { "role": "user", "content": "Hi" } ]
+	}`)
+	s1 := deriveOpenAIContentSessionSeed(compact)
+	s2 := deriveOpenAIContentSessionSeed(spaced)
+	require.Equal(t, s1, s2, "different formatting of identical JSON should produce the same seed")
+}
+
+func TestDeriveOpenAIContentSessionSeed_ResponsesAPI_InputTextTypedItem(t *testing.T) {
+	body := []byte(`{
+		"model": "gpt-5.4",
+		"input": [{"type": "input_text", "text": "Hello world"}]
+	}`)
+	seed := deriveOpenAIContentSessionSeed(body)
+	require.Contains(t, seed, "|first_user=")
+	require.Contains(t, seed, "Hello world")
+}
+
+func TestDeriveOpenAIContentSessionSeed_ResponsesAPI_TypedMessageItem(t *testing.T) {
+	body := []byte(`{
+		"model": "gpt-5.4",
+		"input": [{"type": "message", "role": "user", "content": "Hello from typed message"}]
+	}`)
+	seed := deriveOpenAIContentSessionSeed(body)
+	require.Contains(t, seed, "|first_user=")
+	require.Contains(t, seed, "Hello from typed message")
+}

backend/internal/service/openai_gateway_service.go

@@ -1121,6 +1121,7 @@ func (s *OpenAIGatewayService) ExtractSessionID(c *gin.Context, body []byte) str
 //  1. Header: session_id
 //  2. Header: conversation_id
 //  3. Body:   prompt_cache_key (opencode)
+//  4. Body:   content-based fallback (model + system + tools + first user message)
 func (s *OpenAIGatewayService) GenerateSessionHash(c *gin.Context, body []byte) string {
 	if c == nil {
 		return ""
@@ -1133,6 +1134,9 @@ func (s *OpenAIGatewayService) GenerateSessionHash(c *gin.Context, body []byte)
 	if sessionID == "" && len(body) > 0 {
 		sessionID = strings.TrimSpace(gjson.GetBytes(body, "prompt_cache_key").String())
 	}
+	if sessionID == "" && len(body) > 0 {
+		sessionID = deriveOpenAIContentSessionSeed(body)
+	}
 	if sessionID == "" {
 		return ""
 	}

backend/internal/service/openai_gateway_service_test.go

@@ -237,6 +237,60 @@ func TestOpenAIGatewayService_GenerateSessionHashWithFallback(t *testing.T) {
 	require.Equal(t, "", empty)
 }
 
+func TestOpenAIGatewayService_GenerateSessionHash_ContentFallback(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	rec := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(rec)
+	c.Request = httptest.NewRequest(http.MethodPost, "/openai/v1/chat/completions", nil)
+
+	svc := &OpenAIGatewayService{}
+
+	body := []byte(`{"model":"gpt-5.4","messages":[{"role":"system","content":"You are helpful."},{"role":"user","content":"Hello"}]}`)
+
+	hash := svc.GenerateSessionHash(c, body)
+	require.NotEmpty(t, hash, "content-based fallback should produce a hash")
+
+	hash2 := svc.GenerateSessionHash(c, body)
+	require.Equal(t, hash, hash2, "same content should produce same hash")
+
+	bodyExtended := []byte(`{"model":"gpt-5.4","messages":[{"role":"system","content":"You are helpful."},{"role":"user","content":"Hello"},{"role":"assistant","content":"Hi!"},{"role":"user","content":"How are you?"}]}`)
+	hashExtended := svc.GenerateSessionHash(c, bodyExtended)
+	require.Equal(t, hash, hashExtended, "hash should be stable across later turns")
+
+	bodyDifferent := []byte(`{"model":"gpt-5.4","messages":[{"role":"user","content":"Different question"}]}`)
+	hashDifferent := svc.GenerateSessionHash(c, bodyDifferent)
+	require.NotEqual(t, hash, hashDifferent, "different content should produce different hash")
+}
+
+func TestOpenAIGatewayService_GenerateSessionHash_ExplicitSignalWinsOverContent(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	rec := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(rec)
+	c.Request = httptest.NewRequest(http.MethodPost, "/openai/v1/chat/completions", nil)
+
+	svc := &OpenAIGatewayService{}
+	body := []byte(`{"model":"gpt-5.4","messages":[{"role":"user","content":"Hello"}]}`)
+
+	contentHash := svc.GenerateSessionHash(c, body)
+	require.NotEmpty(t, contentHash)
+
+	c.Request.Header.Set("session_id", "explicit-session")
+	explicitHash := svc.GenerateSessionHash(c, body)
+	require.NotEmpty(t, explicitHash)
+	require.NotEqual(t, contentHash, explicitHash, "explicit session_id should override content fallback")
+}
+
+func TestOpenAIGatewayService_GenerateSessionHash_EmptyBodyStillEmpty(t *testing.T) {
+	gin.SetMode(gin.TestMode)
+	rec := httptest.NewRecorder()
+	c, _ := gin.CreateTestContext(rec)
+	c.Request = httptest.NewRequest(http.MethodPost, "/openai/v1/chat/completions", nil)
+
+	svc := &OpenAIGatewayService{}
+	require.Empty(t, svc.GenerateSessionHash(c, []byte(`{}`)))
+	require.Empty(t, svc.GenerateSessionHash(c, nil))
+}
+
 func (c stubConcurrencyCache) GetAccountWaitingCount(ctx context.Context, accountID int64) (int, error) {
 	if c.waitCounts != nil {
 		if count, ok := c.waitCounts[accountID]; ok {